Synthesis and structural, spectroscopic, and electrochemical characterization of benzo[c]quinolizinium and its 5-aza-, 6-aza, and 5,6-diaza analogues

Article abstract of DOI:10.1016/j.tet.2011.03.023

Four heterocyclic salts 1a-d were prepared by Ca²⁺-assisted cyclization of fluoro derivatives 3, and investigated by spectroscopic (NMR and UV), electrochemical, and computational (DFT and MP2) methods. The mechanism for the formation of the ca

Full text of DOI:10.1016/j.tet.2011.03.023

Tetrahedron 67 (2011) 3317e3327
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Synthesis and structural, spectroscopic, and electrochemical characterization
of benzo[c]quinolizinium and its 5-aza-, 6-aza, and 5,6-diaza analogues
,
b
,
,
b
,
a,
b,
x
Aleksandra Jankowiak a, Emilia Obijalska a y, Piotr Kaszynski a , Adam Pieczonka
,
*
Victor G. Young, Jr. c
a Organic Materials Research Group, Department of Chemistry, Vanderbilt University, Box 1822 Station B, Nashville, TN 37235, USA
b
ꢀ ꢀ
ꢀ ꢀ
Faculty of Chemistry, University of Łodz, Tamka 12, 91403 Łodz, Poland
c Crystallographic Laboratory, Department of Chemistry, University of Minnesota, Twin Cities, MN 55455, USA
a r t i c l e i n f o
a b s t r a c t
Four heterocyclic salts 1aed were prepared by Ca2þ-assisted cyclization of fluoro derivatives 3, and
investigated by spectroscopic (NMR and UV), electrochemical, and computational (DFT and MP2)
methods. The mechanism for the formation of the cations was investigated at the DFT level of theory. 2-D
NMR spectroscopy for 1[ClO4] in DMSO-d6 aided with DFT results permitted the assignment of 1H and
13C NMR signals in cations 1. The molecular and crystal structures for 1a[ClO4] [C13H10ClNO4 triclinic,
Article history:
Received 4 December 2010
Received in revised form 4 March 2011
Accepted 9 March 2011
Available online 16 March 2011
¼67.615(1)ꢁ,
b
ꢁ
¼78.845(2)ꢁ, ¼87.559(2)ꢁ;
g
ꢁ
ꢁ
ꢁ
ꢁ
a
Pꢀ1, a¼9.6517(12) A, b¼11.0470(13) A, c¼12.2373(15) A,
3
ꢁ
V¼1183.0(2) A , Z¼4] and 1d[ClO4] [C12H9ClN2O4 triclinic, Pꢀ1, a¼5.9525(6) A, b¼8.3141(9) A, c¼12.2591
3
ꢁ
ꢁ
(13) A,
a
¼73.487(1)ꢁ,
b
¼83.814(1)ꢁ,
g
¼83.456(1) ; V¼576.07(10) A , Z¼2] were determined by X-ray
ꢁ
crystallography and compared with results of DFT and MP2 calculations. Electrochemical analysis gave
the reduction potential order (1b>1cw1d>1a), which is consistent with computational results.
Ó 2011 Elsevier Ltd. All rights reserved.
1. Introduction
nitrogen atoms, and investigated the effect of the bridge modifica-
tion on the molecular and electronic properties of the cations.
The benzo[c]quinolizinium1,2 (1a) and several of its derivatives3,4
have been known for over four decades. The cation has been in-
tensely investigated as a possible pharmacophore for the treatment
of cystic fibrosis and diseases related to smooth muscle cell con-
striction, such as arterial hypertension and asthma.5 It has also been
used as a structural element of highly solvatochromic dyes4 and
columnar ionic liquid crystals.6 Recently, a diaza analogue of 1a,
triazinium 1b,7,8 and an amino derivative of the mono-aza analogue
1c9 were described in the literature. It was demonstrated that 1b is
highly cytotoxic to cisplatin-resistant tumor cell lines,7 intercalates
to the minor groove of DNA,8 and undergoes a reversible 1eꢀ re-
duction process.8 For these reasons cations, such as 1a and 1b are
attractive for development of pharmacologically active compounds
and molecular materials with multi-redox properties, tunable
electronic absorption spectra, and self-organization. Therefore, we
focused on a series of cations 1aed, in which the CH groups in
the bridging 5,6 positions of 1a are systematically replaced with
N
X
Y
a: X = Y = CH
b: X = Y = N
1
c: X = N, Y = CH
d: X = CH, Y = N
The preparation of cation 1a involves high temperature intra-
molecular cyclization of chlorostilbazole 2a formally by intra-
molecular nucleophilic substitution reaction.1 In contrast, the
preparation of 1b and 1c involves 6
p electrocyclization as the
key step followed by oxidative rearomatization. Thus, the quinoli-
zinium cation 1b was obtained at ambient temperature in Au(III)-
catalyzed light-induced electrocyclization of phenylazopyridine.
Later, the Lewis acid was replaced with a Brønsted acid in prepa-
ration of 1b and several of its derivatives.8 The proposed mecha-
nism involves sunlight trans-to-cis isomerization, acid-induced
tautomerization, electrocyclization, and aerial oxidative rear-
omatization.8 The reaction requires concentrated mineral acid and
long exposures to light, which limits its practicality. The formation
of 9-amino derivative of 1c is a Cu(II)-mediated process in which
* Corresponding author. Tel.: þ1 615 322 3458; e-mail address: piotr.kaszynski@
vanderbilt.edu (P. Kaszynski).
y
A visiting student from the laboratory of Professor Grzegorz Mloston, University
ꢀ
ꢀ
of qodz Poland.
x
ꢀ
ꢀ
In part described in M.Sc. Thesis, University of qodz, Poland, 2008.
0040-4020/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2011.03.023

3318
A. Jankowiak et al. / Tetrahedron 67 (2011) 3317e3327
a radical cation formed by intramolecular one-electron transfer to
the coordinated metal undergoes electrocyclization.9 Cation 1c was
not formed under Brønsted acid conditions described for 1b.9
We previously demonstrated that the azo group activates halogen
toward nucleophilic aromatic substitution.10 Particularly reactive was
fluoride, which undergoes substitution with the thiolate at ambient
temperature. Therefore, we considered fluorophenylazopyridine 3b
as a suitable substrate for a facile intramolecular cyclization to cation
1b inwhich the fluoride is replaced by pyridine. Such a reaction could
offer practical access to functionalized derivatives of 1b under mild
and pH neutral conditions, and could be extended to the preparation
of remaining three cations 1a, 1c, and 1d.
direct sunlight instead of halogen lamp was ineffective and no
product formation was observed after several hours.
The Ca2þ-assisted cyclization proved to be successful also for the
preparation of salts 1c and 1d from appropriate fluorides. Thus,
thermal cyclization of 3c and 3d in hot PhCN in the presence of
solid Ca(OTs)2 gave complete conversion of the fluorides and for-
mation of the cations as the main identifiable products. The re-
action mixture contained substantial amounts of intractable
products presumably formed by decomposition of the azomethine
group labile under the reaction conditions. Cations 1c and 1d were
isolated as perchlorates in 12% and 18% yield, respectively.
The attempted cyclization of 3d in (i) hot PhCN in the absence of
Ca(OTs)2, or (ii) in MeCN with Ca(OTs)2 and irradiated with a halo-
gen lamp gave only the starting material.
Here we describe the preparation of cations 1aed by Ca2þ
-
assisted intramolecular cyclization of fluorides 3. The mechanism of
this transformation is investigated with DFT computational
methods. The effect of substitution of N for a CH in the cations on
their molecular and electronic structures is probed using XRD,
spectroscopic and electrochemical methods, and the experimental
work is augmented with DFT and MP2 level calculations.
The auxiliary role of Ca2þ in the successful preparation of cations
1bed prompted us to investigate cyclization reactions of fluo-
rostilbazole 3a in the presence of Ca(OTs)2. Under reaction con-
ditions developed for the preparation of 1b (MeCN, Ca(OTs)2,
halogen lamp, 10% water) fluoride 3a showed the formation of
cation 1a as the sole product. The reaction time for the complete
transformation of 3a was longer that than for cyclization of 3b,
presumably due to lower absorption in the visible range of the
spectrum. The same reaction of fluoride 3a in the absence of
Ca(OTs)2 showed no product formation.
2. Results and discussion
2.1. Synthesis
Benzo[c]quinolizinium (1a) was prepared according to a modi-
fied literature procedure1 by thermally induced cyclization of
chloro derivative 2a in benzonitrile at 160 ꢁC in the presence of
catalytic amounts of I2, which permitted transecis equilibration.
The product was obtained in 54% yield as the perchlorate salt 1a
[ClO4] (Scheme 1).
Preparation of stilbazoles 2a1 and 3a11 followed literature
procedures and involved acetic anhydride-assisted condensation
of appropriate halobenzaldehyde and 2-picoline (Scheme 2).
Halodiazenes 2b and 3b were prepared by condensation of
2-chloronitrosobenzene12 (4) and 2-fluoronitrosobenzene13 (5),
respectively, with 2-aminopyridine under basic conditions, as
described for the chloro derivative 2b.12 Schiff bases 3c and 3d were
prepared using standard acid-catalyzed condensation of appropri-
ate aldehyde and amine (Scheme 2).
Hal
X
Y
R1
R2
N
Hal
2-E, Hal = Cl
2, Hal = Cl
3, Hal = F
N
+
3-E, Hal = F
a: R1 = CHO, R2 = CH3; Ac2O
b: R1 = NO, R2 = NH2; base
N
c: R1 = NH2, R2 = CHO; acid cat.
d: R1 = CHO, R2 = NH2; acid cat.
1
X
Y
Scheme 2.
Scheme 1.
2.2. Molecular and crystal structures
Similar thermal reaction of chloro diazene 2b or fluoro diazene
3b in benzonitrile in the absence of I2, or in warm MeCN using
sunlight or strong Pyrex-filtered halogen lamp light to effect
transecis isomerization did not lead to the formation of the tri-
azinium ion 1b. No reaction of 2b was observed either in refluxing
MeCN nor in hot PhCN. Similar reactions of fluoro derivative 3b
were equally fruitless even though the fluoride usually is a signifi-
cantly more mobile leaving group than chloride.
To activate the halogen toward displacement and intra-
molecular cyclization, we used an equimolar amount of AgOTs as
a strong halophilic reagent. Reaction of chloro derivative 2b con-
ducted in warm MeCN irradiated with a halogen lamp led to the
formation of 1b in about 10% yield before the reaction was stopped
by a silver mirror formed on the vessel’s walls.
A reaction of the fluoro analog 3b in warm MeCN in the presence
of fluorophilic Ca2þ ions led to complete conversion of the substrate
to the cation 1b, isolated in w80% yield as the tosylate 1b[OTs]. The
reaction mixture contained about 10% of water to ensure solubility
of Ca(OTs)2. No product formation was observed when the reaction
was run without Ca(OTs)2. Exposure of the reaction mixture to
Crystals of 1a[ClO4] and 1d[ClO4] were grown by slow evapo-
ration of MeCN solutions. Solid-state structures for both com-
pounds (Fig. 1) were determined by X-ray diffraction14 and their
selected bond lengths are shown in Table 1.
The two structures are essentially planar. The C(1)eN(11)eC
(10a)eC(10) angle in cation 1a is 2.1(2)ꢁ in molecule A and 5.6(2)ꢁ
in molecule B. In the structure 1d, the deviation from planarity is
similar [2.4(3)ꢁ]. The cations are characterized by a short C(10a)eN
ꢁ
(11) distance of 1.425(2) A in both structures. This distance is
comparable to that reported for 9-amino derivative of 1c (1.424
9
8
ꢁ
ꢁ
(3) A), longer than that observed in 1b (1.401(5) A), and signifi-
cantly shorter than the analogous PheN distance in N-phenyl-
pyridinium15 (1.45 A). The short NePh distance is a consequence of
ꢁ
the decreasing distance between the bridging atoms X(6)eY(5)
ꢁ
ꢁ
from 1.342(3) A in 1a through 1.297(3) A in 1d to the shortest, 1.275
ꢁ
(6) A in 1b. With the falling XeY distance, the C(1)/C(10) bay in
the cations opens up by about 0.06 A in 1b in comparison to 1a
ꢁ
(Table 1). These results are well reproduced by computational
methods (vide infra).

A. Jankowiak et al. / Tetrahedron 67 (2011) 3317e3327
3319
B3LYP and MP2(fc) methods and several bases sets, and results were
compared to experimental solid-state structures of 1a, 1b, and 1d.16
In general, all fourcationsconvergedto planarstructuresusing the
DFT method with four basis sets: 6-31G(d,p), 6-311G(2d,p), 6-311þG
(2d,p), and cc-pVDZ, and the MP2 level of theory with 6-31G(d,p), 6-
311G(2d,p), and cc-pVDZ basis sets. The only exception is cation 1a,
which is predicted by the MP2/6-31G(d,p) method to be twisted with
the C(1)eN(11)eC(10a)eC(10) torsional angle of 6.7ꢁ.
Detailed comparison of theoretical and experimental bond
lengths demonstrated that the DFT method performed better than
the MP2 method in geometry optimization of cations 1.16 In both
methods the cc-pVDZ basis set performed least satisfactory giving
the largest mean error. Results closest to the experimental in-
teratomic distances were obtained with the B3LYP/6-311G(2d,p)
method, and selected data are shown in Table 1. For cations 1a and
Fig. 1. Thermal ellipsoid diagram for 1a[ClO4] (molecule A) and 1d[ClO4] drawn at 50%
probability. For clarity the anion is omitted. Interatomic distances are listed in Table 1.
Atom numbering scheme according to the chemical structure.
ꢁ
1d the mean difference was 0.001ꢂ0.005 A (absolute mean
ꢁ
0.004ꢂ0.005 A), which is close to the typical experimental un-
Table 1
ꢁ
certainty of ꢂ0.003 A. The largest difference between the calculated
Selected experimental and calculated bond lengths for salts 1[ClO4]
and experimental distances was found for the C(5)eC(6) bond in 1a
and C(4a)eN(5) in 1d. With these two results excluded from
9
10
1
2
10a 11
ꢁ
8
3
analysis the error decreases to 0.002ꢂ0.004 A (absolute mean
N
ꢁ
0.003ꢂ0.002 A). The same analysis for cation 1b gave higher mean
7
6a
4a
4
ꢁ
ꢁ
X
Y
error 0.005ꢂ0.010 A (absolute mean 0.010ꢂ0.005 A) for all bonds,
6
5
which is consistent with significant experimental uncertainty of
ꢁ
ꢂ0.006 A (Table 1).
ꢁ
Distances/A
1a
1b
Calcdb Exptlc
1c
1d
On the basis of geometry optimization results, the B3LYP/6-311G
(2d,p) methodwas usedin subsequentmechanisticstudies (Table 2),
assignment of NMR chemical shifts (Table 3), analysis of electronic
absorption spectra (Table 4), and investigation of reduction for cat-
ions 1 (Table 5).
Exptla
Calcdb Calcdb Exptla
Calcd
C(4)eC(4a)
C(4a)eY(5)
Y(5)eX(6)
X(6)eC(6a)
1.402(2) 1.408 1.379(6) 1.400 1.400 1.397(3) 1.405
1.426(2)d 1.418d 1.387(5)e 1.376e 1.432f 1.372(2)g 1.354g
1.342(3)d 1.353d 1.275(6)e 1.273e 1.290f 1.297(3)g 1.298g
1.429(2)d 1.423d 1.387(5)e 1.377e 1.372f 1.422(3)g 1.425g
C(10a)eC(6a) 1.408(2) 1.415 1.401(5) 1.411 1.417 1.403(3) 1.410
C(10a)eN(11) 1.425(2) 1.426 1.401(5) 1.413 1.421 1.425(2) 1.422
Table 2
C(1)eN(11)
1.379(2) 1.376 1.375(4) 1.369 1.370 1.376(2) 1.374
Calculated relative free energy (kcal/mol) for compounds involved in conversion of
chlorides 2 to 1 and fluorides 3 to 6 in MeCNa
C(4a)eN(11) 1.384(2) 1.385 1.385(4) 1.377 1.378 1.384(2) 1.388
C(1)/C(10)
2.863
2.874 2.91
2.936 2.903 2.889
2.903
c
d
D
G298
D
E/Z
Gz
D
Z/E
Gz
D
Gz
DG298
298
298
298
a
ZdEb
Z-I/1-Cl
Z-I/6e
Z-Ie1-Cl
Distances for molecule A. See Ref. 14.
B3LYP/6-311G(2d,p) level geometry optimization at Cs point group symmetry.
b
Z-Ie6e
Numbering system according to the chemical structure.
c
CF3SO3ꢀ salt, Ref. 8.
a
b
c
Cl
F
Cl
F
Cl
F
Cl
F
8.3
10.6
15.8
17.2
6.2
8.2
10.2
11.0
>45
>45
40.2
35.3
20.0
19.9
23.6
18.8
>35
>35
24.5
17.7
13.7
11.6
13.4
7.8
31.4
24.9
29.9
22.4
35.0
27.3
32.0
24.2
ꢀ23.7
16.3
d
X¼Y¼CH.
e
ꢀ17.8
X¼Y¼N.
f
5.7
X¼N, Y¼CH.
g
ꢀ15.5
17.7
X¼CH, Y¼N.
d
ꢀ20.6
The unit cell of 1a[ClO4] contains two distinct cations and two
anions per asymmetric unit. These cations and anions are arranged
about an approximate non-crystallographic local center at ap-
proximately (0.22, 0.73, 0.30). This serves only to increase the
contents of the asymmetric unit to Z0¼2. Both cations are disor-
dered leading to exchange of the C(10a) and N(11) positions, each in
different ratios: 0.95:0.05 in cation A and 0.72:0.28 in cation B. The
cations are aligned approximately parallel to each other with the
12.6
a
B3LYP/6-311þG(2d,p)//B3LYP/6-311G(2d,p) level calculations in MeCN di-
electric medium.
b
D
D
D
G298¼G298(Z)ꢀG298(E) for the two global minima conformers Z and E.
c
Gz
for conversion of the global minimum conformer of E to Z.
Gz298 for conversion of the global minimum conformer of Z to E.
d
e
298
Upper line
D
G298¼G298(Z-I)ꢀG298(1-Cl). Clꢀ coordinated to 1. Lower line,
DG298¼G298(Z-I)ꢀG298(6).
ꢁ
separation between them of about 3.35 A.
In the unit cell of 1d[ClO4] there are two cations and two anions
related by an inversion center. The cations form an infinite slipped
stack allowing for partial overlap of the pyridinium rings separated
2.4. Mechanistic analysis
The mechanism for the formation of cations 1 from halo pyri-
dine derivatives 2 and 3 involves E to Z isomerization and sub-
sequent intramolecular attack of the pyridine N atom on the C(Hal)
atom of the halophenyl ring. In principle, the cation 1 can be formed
directly from the cis isomer or through non-aromatic cyclic product
6 (Scheme 3), which undergoes rearomatization upon departure of
the halide. The success of the reaction depends on (i) the access to
the Z isomer, and (ii) relative height of the barriers to cyclization
versus back isomerization to the E isomer. To assess these factors,
transformations of chloro and fluoro pyridine derivatives 2 and 3 to
the cations 1 were investigated at the B3LYP/6-311þG(2d,p)//
B3LYP/6-311G(2d,p) level of theory in MeCN dielectric medium
without participation of metal ions.
ꢁ
by about 3.4 A within the unit cell, and partial overlap of benzene
ꢁ
rings between the unit cells separated by about 3.3 A. Cations in the
neighboring unit cells are nearly co-planar, and they are separated
ꢁ
by C(4)/N(5) distance of 3.53 A. Each pair is separated from the
next by two anions.
2.3. Geometry optimization and selection of a computational
model
For a better understanding of properties of cations 1, we con-
ducted quantum-mechanical calculations. To select a suitable com-
putational protocol, equilibrium geometries of 1 were obtained with

3320
A. Jankowiak et al. / Tetrahedron 67 (2011) 3317e3327
Table 3
Assignment of experimental and calculated 13C and 1H NMR chemical shifts for 1[ClO4]a
9
10
1
2
4
10a 11
8
3
N
7
6a
4a
X
Y
5
6
Nucleus
1ab
1bc
1cd
1de
13C
1H
13C
1H
13C
1H
13C
1H
C(1)eH
C(2)eH
C(3)eH
C(4)eH
C(4a)
134.6
131.4
124.3
124.3
140.6
140.9
128.5
129.5
143.3
143.7
123.0
123.1
136.6
139.7
126.8
127.7
130.5
132.1
130.5
131.6
132.9
134.7
118.1
115.3
134.3
135.5
10.40
9.80
8.29
8.02
8.65
8.40
8.70
8.37
d
133.0
129.9
129.2
130.5
146.0
147.0
130.5
133.4
141.4
141.4
d
10.51
9.77
8.74
8.45
9.18
9.00
9.46
9.27
d
134.9
131.6
128.7
129.6
143.7
144.2
129.1
130.6
134.4
132.8
150.6
150.5
d
10.54
9.91
8.64
8.36
8.99
8.76
9.01
8.67
d
134.1
130.7
124.5
125.1
145.2
145.0
131.4
131.2
147.6
148.3
d
10.31
9.68
8.36
8.05
8.83
8.64
8.68
8.54
d
Y(5)
8.38
8.07
8.74
8.54
d
d
d
d
9.88
9.59
d
d
X(6)
d
162.4
164.0
121.9
121.4
130.8
132.9
131.5
133.0
137.6
140.4
117.3
115.0
135.9
137.3
10.00
9.66
d
C(6a)
142.4
144.0
132.3
136.4
133.3
134.9
138.9
142.8
116.8
114.0
123.1
122.3
139.2
141.9
131.2
134.1
132.4
133.9
132.5
135.3
117.7
115.0
126.7
127.2
d
C(7)eH
C(8)eH
C(9)eH
C(10)eH
C(10a)
8.40
8.30
8.08
8.05
8.21
8.23
9.16
8.83
d
9.13
9.18
8.42
8.45
8.60
8.59
9.23
8.76
d
8.44
8.55
8.18
8.21
8.22
8.24
9.15
8.80
d
8.63
8.47
8.23
8.20
8.49
8.48
9.17
8.80
d
a
Upper experimental data in DMSO-d6, lower (italics) data derived from B3LYP/6-311þG(2d,p)//B3LYP/6-311G(2d,p) level calculations in DMSO dielectric medium. For
details see Supplementary data.
b
X¼Y¼CH.
c
X¼Y¼N.
d
X¼N, Y¼CH.
e
X¼CH, Y¼N.
Table 4
Table 5
Selected experimentala and calculated b electronic transition energies and oscillator
Experimental and calculated electrochemical reduction potentials for 1[ClO4] versus
strength for 1[ClO4]
SCE
Compound
Experimentala
Theoreticalb
(f )
Compound
Experimentala
Calculatedb
1 red
1 red c
Absorption
lmax/nm
Emission
lem lex)/nm
p
/p
*
E1/2
E
E1 red
(þ/ꢃ)/V
E2 red
pc
(
(þ/ꢃ)/V
(þ/ꢃ)/V
(ꢃ/ꢀ)/V
d
1a[ClO4]
1b[ClO4]
1c[ClO4]
1d[ClO4]
363,c 257d
359,c 259f
363,c 242
349,c 250
394 (347)e
dg
394 (347)e
d
341 (0.21), 281 (0.29), 259 (0.34)
340 (0.23), 268 (0.51)
347 (0.18), 286 (0.13), 268 (0.34)
320 (0.15), 273 (0.19), 261 (0.45)
1a
1b
1c
1d
ꢀ0.91
ꢀ0.24
ꢀ0.66
ꢀ0.67
ꢀ0.78
ꢀ1.84
ꢀ1.28
ꢀ1.64
ꢀ1.64
ꢀ0.21e,f
0.21
d
ꢀ0.28
ꢀ0.57
d
a
Recorded in MeCN.
b
Obtained with TD B3LYP/6-311þ(2d,p)//B3LYP/6-311(2d,p) method in MeCN
dielectric medium.
c
The (0,0) transition.
365 nm (log
Excitation wavelength.
Similar spectral characteristic reported in Ref. 8.
ꢀ0.52g
ꢀ0.32
ꢀ1.48
h
Me
N
CN
d
e¼3.68) and 255 nm (log
e
¼4.07) in EtOH. Ref. 3.
e
f
a
Recorded in MeCN (0.1 M [Bu4N]þ[ClO4]ꢀ) and reported relative to SCE.
B3LYP/6-311þG(2d,p)//B3LYP/6-311G(2d,p) level as
g
Emission at 535 nm (lex¼340 nm) in MeCN reported in Ref. 8.
b
DH in MeCN dielectric
medium.
Initial computational effort concentrated on identifying global
c
Cathodic peak potential.
d
e
f
conformational minima for trans and cis isomers of 2 and 3 by
comparison of SCF energies obtained in MeCN medium. Results for
2-E and 3-E showed that the most stable conformation for a, b, and
d in MeCN is of type E-IV, while conformer E-III is the most stable
form of 2c-E and 3c-E (Fig. 2). For the cis isomer the most stable
conformers in MeCN are of type Z-II (2c and 3c), Z-III (2a, 2b, 3a),
Irreversible process.
Second cathodic wave at Epc
2 red
¼ꢀ0.89 V.
2 red
1 red
lit. E1/2
¼ꢀ0.26 V and Epc
¼ꢀ1.10 V (0.1 M [Et4N]þ[ClO4]ꢀ in MeCN versus
SCE) Ref. 8.
g
[Bu4N]þ[ClO4]ꢀ (0.25 M) in MeCN versus Ag/Agþ (0.01 M in MeCN) assumed to
be 0.44 V versus SCE. Ref. 26.
h
Not reported.

A. Jankowiak et al. / Tetrahedron 67 (2011) 3317e3327
3321
Hal
but it was estimated at >40 kcal/mol. Analysis of the TS structures
demonstrates that the cisetrans isomerization for the nitrogen-
containing derivatives 2bed and 3bed proceeds through inversion
at the N atom, which is consistent with previous computational
results for phenylazopyridine17 and N-benzylideneaniline.18e20
Results shown in Table 2 demonstrate that cis and trans forms
interconvert thermally only for azomethine derivatives c and d.
Hal
N
X
Y
N
X
Y
E
Z
The barrier D
Gz298 calculated for 3c and 3d is about 19 kcal/mol for
the E/Z isomerization and is less than 12 kcal/mol for the reverse
process. The analogous barriers to isomerization for the chloro
derivatives 2c and 2d are higher by up to 5.6 kcal/mol calculated
for 2d-Z/2d-E. For all four compounds the concentration of the
cis isomers at the equilibrium is small, less than 0.1%. The calcu-
lated low activation energy for Z/E conversion is consistent with
that experimentally established for isomerization of benzylidea-
niline (Ea¼17.3 kcal/mol in MeCN)21 and pyridyl phenyl azome-
thine (Ea¼13.1e14.3 kcal/mol),22 which are the closest
experimental models of c and d.
Hal
N
N
X
Y
X
Y
1
6
Scheme 3.
and Z-IV (2d, 3d). This preference for conformational minimum is
different in gas phase. Selected optimized structures involved in the
transformation of 2b and 3b to 1b are shown in Fig. 3.
Hal
Hal
X
X
X
X
N
N
Y
Y
Y
Y
Hal
Hal
N
N
E-I
E-II
E-III
E-IV
Hal
Hal
N
N
X
Y
X
Y
X
Y
X Y
N
N
Hal
Hal
Z-I
Z-II
Z-III
Z-IV
Fig. 2. Conformers of E and Z isomers.
Fig. 3. B3LYP/6-311G(2d,p) optimized geometries for selected structures involved in cyclization 3b to 6b and 1-Cl formed from 2-Z-I.
All four conformers of 3a-E, and also 2a-E-I, 2b-E-III, 2d-E-IV,
and 3b-E-III converged to planar structures as confirmed by fre-
quency calculations for Cs-constrained molecules. Surprisingly,
fully optimized structure of one conformer of cis stilbazole, 3a-Z-IV,
also has the Cs symmetry, apparently due to favorable CeH/N
Results for azenes b indicate that the cis isomers 2b-Z and 3b-Z
cannot be thermally populated due to high E/Z isomerization bar-
rier of
Gz298 >35 kcal/mol, but both cis forms can thermally isom-
D
erize to the more stable trans isomers 2b-E and 3b-E. The calculated
height of the barrier of
the 2b-Z/2b-E process is in good agreement with the experimental
D
Gz ¼24.5 kcal/mol (Table 2) in MeCN for
298
ꢁ
interactions (d¼2.102 A).
The energy difference between the cis and trans forms was
calculated for the global minima of each isomer. For simplicity, the
activation energy for phenylazopyridine (
D
Gz ¼25.5 kcal/mol,
298
D
Sz¼ꢀ14.5 cal/molK).23 The same barrier for the fluoro analogue is
free energy of activation
D
Gz
for the cis/trans isomerization
lower by nearly 7 kcal/mol. Finally, high barriers to isomerization
estimated for stilbazoles 2a and 3a and established experimentally
for Z-stilbene (Ea¼42.8 kcal/mol)24 prevent thermal isomerization of
either a-Z or a-E.
298
was calculated for 2-Z and 3-Z global minimum conformers in
MeCN (Table 2). The transition state structure for isomerization of
stilbazoles 2a and 3a could not be located using standard methods,

3322
A. Jankowiak et al. / Tetrahedron 67 (2011) 3317e3327
The intramolecular cyclization of halophenyl derivatives occurs
form the 2-Z-I and 3-Z-I conformers and proceeds by an apparent
electrocyclic process. The former yields nearly planar cations
with the chloride anion weakly coordinated to the C1 position
Computational results also helped with the assignment of the
resonance to the C(5) and C(6) nuclei in 1a. Assigned experimental
and calculated chemical shifts are shown in Table 3.
A comparison of the computed chemical shifts with experi-
mental values in Table 3 demonstrates a close agreement for the 13C
nuclei. The mean difference between experimental and theoretical
values is ꢀ0.6ꢂ2.0 ppm (absolute 1.7ꢂ1.1 ppm). The largest differ-
ences Dd between experiment and theory are observed for the C(1)
6
p
ꢁ
ꢁ
(dC1Cl¼2.3 Ae2.5 A) in 1-Cl, while the cyclization of the later, flu-
orophenyl derivatives 3-Z-I, leads to non-aromatic product 6. The
free energy of activation
D
Gz
required for the cyclization of
298
chlorophenyls 2-Z-I is in a range of about 30 kcal/mol (2b-Z)e
35 kcal/mol (2c-Z) and the reaction is exergonic by 18e24 kcal/mol.
In contrast, cyclization of fluorophenyl 2-Z-I to the non-aromatic 6
has lower activation barrier by 6e8 kcal/mol than that for the
chloro analogues, but the process is significantly endergonic by
6e18 kcal/mol and thermally reversible. The least unfavorable cy-
carbon in all cations
(
(Dd >3 ppm), and for C(7) in 1b
Dd¼ꢀ4.1 ppm). 1H NMR chemical shifts are less accurately pre-
dicted and larger differences Dd are observed especially for the C
(1)eH nuclei for which Dd values are in a range of 0.59e0.74 ppm.
The mean deviation for all 1H NMR datapoints is 0.23ꢂ0.21 ppm
(absolute 0.25ꢂ0.19 ppm), while without the C(1)eH nuclei it im-
proves to 0.18ꢂ0.16 ppm (absolute 0.20ꢂ0.13 ppm).
clization is calculated for 3b-Z (
D
G298¼5.7 kcal/mol), while the
formation of 6a and 6c is most endergonic and requires additional
10.6 and 12 kcal/mol, respectively. This order of energies is con-
sistent with that of the CeF distance in 6: the shortest is found in 6b
2.6. Electronic spectra
ꢁ
ꢁ
(dCF¼1.446 A) and the longest in 6a (dCF¼1.471 A).
The difference in behavior of the two halides presumably results
from much better solvation of the chloride in organic media than
that of Fꢀ. As a consequence, the progress of the reaction of 3-Z is
‘frozen’ at the stage of 6, while reaction of chlorophenyl derivatives
2 progresses to the formation of cation 1.
Experimental absorption and emission spectra for salts 1a
[ClO4]e1d[ClO4] were recorded in MeCN and selected data are
presented in Fig. 4 and Table 4.
2.5 104
2 104
1.5 104
1 104
5 103
0
Computational data are consistent with experimental results,
which demonstrate that only 1a-Cl can be obtained thermally from
2a in a metal-ion-free process. The cyclization reaction is exo-
thermic and is faster than isomerization of 2a-Z to 2a-E. The later is
also true for 3a-Z but the reaction is significantly endothermic and
the product is not formed. For the remaining chlorophenyl and
fluorophenyl derivatives, the cis-to-trans isomerization reaction is
faster than cyclization and no product formation is observed. In
addition, cyclization reactions of fluorophenyl derivatives 2 are
endothermic.
Experimental results of successful formation of cations 1 upon
addition of Agþ to chlorides 2 and Ca2þ to fluorides 3 suggest that
the ions coordinate to the halogens and significantly alter the po-
tential energy surface for the cyclization and presumably for
isomerization processes. The activation energy for cyclization of the
coordinated halide becomes lower than that for isomerization, and
consequently formation of 1 is faster for halides 2 and 3.
The formation of the requisite cis isomers of the Schiff bases,
3c-Z and 3d-Z, is accomplished thermally, which, in accordance
with the computational results, shows that the two isomers are in
thermal equilibrium. The generation of 2b-Z and 3b-Z isomers in
concentrations sufficient for efficient formation cyclization requires
intense light, since both thermally convert to 2b-E and 3b-E and
neither can be transformed thermally back to the cis forms. Finally,
the preparation of 1a requires I2 catalysis at high temperature to
equilibrate the a-E and a-Z forms. Without the catalyst the requisite
2a-Z and 3a-Z isomers are thermally inaccessible from the trans
forms. However, both can be obtained photochemically.1
200
250
300
350
400
450
500
550
Fig. 4. Absorption (black) and emission (red, lex¼347 nm) spectra for 1a[ClO4]
recorded in MeCN.
All four cations exhibit moderate absorption in the region of
about 340 nm and significantly stronger absorption at about
260 nm. The low energy absorption exhibits vibronic structure with
the splitting of about 1330 cmꢀ1 for 1a and 1d and about 950 cmꢀ1
for 1b. Table 4 lists the (0,0) transitions for the cations. Two of the
four cations fluoresce at ambient temperature. Thus, solutions of 1a
[ClO4] and 1c[ClO4] in MeCN exhibit a moderate fluorescence with
a maximum at 394 nm (lex¼347 nm). Despite efforts, fluorescence
that was reported8 for 1b at 535 nm was not observed even for
a wide range of excitation wavelengths.
To shed more light on the nature and origin of the observed
transitions, we conducted TD DFT calculation in MeCN dielectric
medium. Calculations demonstrated that there are three main
2.5. NMR chemical shift assignment
pep excitations of moderate intensity in a range of about 340 nm,
*
270 nm, and 250 nm. This is consistent with experimental spectra
assuming that the two high energy excitations are merged into
a broad absorption band observed e.g., at 257 nm for 1a (Fig. 1). The
lowest energy excitation involves transition mainly from the HOMO
to the LUMO, both orbitals delocalized over the entire molecule as
shown for 1d in Fig. 5. The higher energy excitation involves
HOMO/LUMOþ1 and HOMO-1/LUMO as the main transitions in
1a and 1d. The excitation in 1b at 286 nm involves HOMO-
2/LUMO, since HOMO-1 is A0 symmetry orbital and contains the
lone pairs. In cation 1c, this medium energy excitation is apparently
split into two bands: one at 329 nm involving the HOMO-1/LUMO
transition, and the second at 286 nm originating from the HOMO-
Structural assignment of 1H and 13C signals in cations 1 was
accomplished by taking advantage of peak multiplicity, by analysis
of 2-D (1He1H and 1He13C) and DEPT-90 NMR spectra recorded in
DMSO-d6 solutions,25 and by using DFT-calculated 13C chemical
shifts in the solvent’s dielectric medium. In all four cations the
assignment started from the highly deshielded proton appearing as
a doublet at the C(1) carbon atom adjacent to the pyridinium
nitrogen atom. Assignment for the benzene ring was helped by DFT
results showing that the C(10) carbon atom is the most shielded in
the molecule and appears at about 115 ppm in the spectrum. The
quaternary carbon atoms, C(4a), C(6a), and C(10a), were identified
by DEPT-90 technique and assigned according to the DFT results.
/LUMOþ1 transition. The next highest
pep excitation at about
*

A. Jankowiak et al. / Tetrahedron 67 (2011) 3317e3327
3323
250 nm involves the HOMO-1/LUMOþ1 transition for 1a, 1c, and
1d, and HOMO-2/LUMOþ1 transition for 1b.
to anions 1L (ꢃ/ꢀ) are calculated to be over 1 V more cathodic. A
comparison to the pyridinium derivatives demonstrates that 1b is
easier to reduce than 4-cyano-N-methylpyridinium26 by about
0.3 V, which in turn is about 0.15 V more anodic than reduction of
1c and 1d.
Analysis of radicals 1ꢃ demonstrates that spin is localized mainly
on the central ring with the most significant spin concentration on
the X(6) atom, while relatively little spin is delocalized onto the
benzene ring (Fig. 6). This is consistent with the major resonance
forms of the radicals shown in Scheme 4. A relatively high positive
spin density is localized at the C(3) position and suggests a poten-
tial substitution site that may help to stabilize the radical. Radicals
1bꢃ and 1cꢃ in which the spin is stabilized by the N(6) atom are
planar. All anions are found to be non-planar.
Fig. 5. B3LYP/6-311þG(2d,p) derived contours and energies of molecular orbitals rel-
evant to low energy excitations for 1d in MeCN dielectric medium.
The non-bonding electrons of the nitrogen atoms constitute the
HOMO-1 in 1b and HOMO-2 in 1c and 1d (Fig. 5). The forbidden
ne
consistent with the observed low intensity absorption band tailing
to about 450 nm. The ne excitations in 1c and 1d are calculated
at 311 nm and 288 nm, respectively, and presumably are obscured
by the more intense bands.
p
*
excitation for 1b is predicted at 429 nm and 310 nm, which is
Fig. 6. Calculated spin density for radical 1ꢃ at the B3LYP/6-311þG(2d,p)//B3LYP/6-
311G(2d,p) level of theory in MeCN dielectric medium. Dark circles represent positive
values. For details see Supplementary data.
p
*
pep
*
A comparison of computational results for the two isomeric
species c and d demonstrates that cation 1d is more thermody-
namically stable than 1c by 2.9 kcal/mol, while radical 1cꢃ is more
stable than 1dꢃ by 3.4 kcal/mol in MeCN dielectric medium. The
former reflects better stabilization of the positive charge by the C(6)
carbon atom delocalized from the pyridinium, while the unpaired
electron in 1ꢃ is better accommodated by the nitrogen atom in the
same position (Scheme 4).
2.7. Reduction of cations 1
One electron reduction of cations 1 leads to p-delocalized het-
eroaromatic radicals 1ꢃ (Scheme 4), which are analogues of pyridyl
radicals obtained from pyridinium salts.26 Further reduction of the
radicals results in anions 1L (Scheme 4). The formation and stability
of these species was studied electrochemicallyand computationally.
3. Conclusions
N
N
Results demonstrate that cations 1aed are accessible from flu-
oro derivatives 3 by Ca2þ-promoted cyclization. The process in-
volves trans-to-cis isomerization of 3 followed by Ca2þ-assisted
electrocyclization to cation 1. In the absence of Ca2þ ions cyclization
of 3-Z leads to an endothermic intermediate 6 through a barrier
higher (for bed) than that for isomerization to 3-E, according to
DFT calculations. The reaction conditions, such as the use of heat or
light to populate the less thermodynamically stable cis isomer 3-Z
necessary for the cyclization, are consistent with the calculated
barriers for isomerization.
X
Y
X
Y
1
-e-
+e-
N
N
X
Y
X
Y
1
-e-
+e-
A comparison of the experimental and theoretical molecular
structures for three of the four cations demonstrated that the B3LYP
method with the 6-311G(2d,p) basis set particularly well re-
produces the experimental geometry and provides a tool for
assessing formation and properties of this class of organic cations.
The DFT method was used successfully to assign NMR chemical
shifts, analysis of electronic absorption spectra, and study of re-
duction of cations 1 to radicals 1ꢃ. In general, experimental results
were well reproduced by DFT calculations.
Spectroscopic analysis demonstrated that all four cations exhibit
two main regions of absorption at about 350 nm and 260 nm re-
lated to pep excitation. Among the cations only 1a and 1c exhibit
*
fluorescence with a maximum at about 394 nm. Fluorescence
reported for 1b was not confirmed.
Electrochemical analysis revealed that only 1b exhibits re-
versible (þ/ꢃ) reduction with half potential of ꢀ0.21 V, while the
remaining three cations undergo irreversible reduction at more
N
1-
X
Y
Scheme 4.
Cyclic voltammetry analysis demonstrated that only 1b exhibits
reversible reduction wave at E1/2¼ꢀ0.21 V and irreversible cathodic
wave at ꢀ0.89 V (Table 5). The remaining three cations undergo
irreversible þ/ꢃ reduction at lower potentials; 0.4 V lower for cat-
ions containing one bridging nitrogen, 1c and 1d, and nearly 0.7 V
lower for 1a. The experimental order of reduction potentials,
1b>1dw1c>1a, is similar to that established computationally
1b>1c>1d>1a (Table 5) and reflects the ability to stabilize the
positive charge in the cations. Potentials for reduction of radicals 1ꢃ

3324
A. Jankowiak et al. / Tetrahedron 67 (2011) 3317e3327
cathodic potentials. Substitution at high spin density site C(3) with
an electron-withdrawing group may lead to stabilization of the
radicals and reversible reduction of the cations.
5. Experimental section
5.1. General remarks
The sequential introduction of the nitrogen atoms to the X(6)eY
(5) bridge of 1a has a significant effect on electrochemical behavior
of the cation, but much less on their photophysical properties. In-
terestingly, the position of the nitrogen atom in the bridge has ef-
fect on photoluminescence properties, relative stability, and the
energy profile for the formation of the two isomeric cations from 3c
and 3d.
Overall, the results demonstrate a potentially general method
for the preparation of heterocyclic cations, analogues of 1aed, and
derivatives that are suitable for incorporation into more complex
molecular architectures. The developed computational protocol
permits the design of new cations and evaluation of their formation
and properties.
Melting points are uncorrected. NMR spectra were recorded at
400 or 500 MHz (1H), and 75 or 100 MHz (13C), respectively, in
CDCl3 (7.26 and 77.0 ppm), CD3CN (1.95 ppm), anhydrous DMSO-d6
(2.49 and 39.5 ppm) as specified. DMSO-d6 was taken from freshly
opened ampule. Chemical shifts were referenced to TMS (1H) or
solvent (13C). UV spectra were recorded in spectroscopic-grade
CH3CN. Molar extinction coefficients were obtained by fitting
maximum absorbance against concentration in agreement with
Beer’s law.
5.2. Photocyclization
Solutions of halide 2 or 3 were placed in a Pyrex NMR tube or
flask and irradiated without filter with a Husky work light
(WL500SPC-H) setup equipped with a 500 W (125e130 V) clear
halogen double ended lamp (T3 base) at a distance of about 50 cm.
The reaction mixtures were spontaneously heated up by the lamp.
4. Computational details
Quantum-mechanical calculations were carried out with the
B3LYP27,28 and MP2(fc)29 methods with 6-31G(d,p), 6-311G(2d,p),
6-31þG(2d,p), or cc-pVDZ basis set using the Gaussian 03 and
Gaussian 09 packages.30,31 Geometry optimizations were un-
dertaken using appropriate symmetry constraints and tight con-
vergence limits. Vibrational frequencies were used to characterize
the nature of the stationary points achieved with the DFT method,
and to obtain thermodynamic parameters. Zero-point energy (ZPE)
corrections were scaled by 0.9806.32 Geometry optimizations at the
MP2 level were initially performed without symmetry constraints,
and the resulting planar structures were reoptimized with the
imposed Cs symmetry.
Electronic excitation energies for cations 1 were obtained at the
B3LYP/6-311þG(2d,p)//B3LYP/6-311G(2d,p) level using the time-
dependent DFT calculations33 supplied in the Gaussian package.
Solvent effect on electronic excitations was included using the PCM
model34 [keywords: SCRF(PCM, Solvent¼CH3CN)].
5.3. Cyclic voltammetry
Electrochemical analysis of 1[ClO4] was conducted under dry
argon in MeCN (distilled over CaH2) containing freshly dried
[Bu4N]þ[ClO4]ꢀ (50 ꢁC, P2O5, vacuum, overnight) as the supporting
electrolyte (0.1 M) and Pt microdisk working electrode with Ag/
AgCl pseudoelectrode as reference. The concentration of 1[ClO4]
was about 1 mM and the scanning rate was 0.1 V/s. Peak potentials
were referenced to the Fc/Fcþ couple by adding small amounts of
ferrocene to the solution, which was assumed to be þ0.462 V ver-
sus SCE (MeCN, 0.1 M [Bu4N]þ[ClO4]ꢀ).40
5.3.1. Benzo[c]quinolizinium perchlorate (1a[ClO4])1. A mixture of
2a (250 mg, 1.1 mmol) and catalytic amounts of I2 was refluxed in
dry benzonitrile (5 mL) for 36 h. The solvent was removed in vac-
uum, the residue was dissolved in water, insoluble brown solid was
filtered, and 70% aqueous HClO4 was added to the filtrate. The
resulting pale yellow solid was filtered, washed with water and
Et2O, and recrystallized (EtOH) to afford 150 mg (54% yield) of 1a
[ClO4] as yellowish crystals: mp 195e197 ꢁC (lit.1 mp 187e189 ꢁC);
NMR shielding tensors for 1 were obtained at the B3LYP/6-
311þG(2d,p)//B3LYP/6-311G(2d,p) level using the GIAO method35
and solvent effects were included using keywords: SCRF(PCM,
Solvent¼DMSO). The computed isotropic shielding parameters for
1 were referenced using chemical shift of benzene calculated at the
same level of theory and experimentally observed36 in DMSO-d6
(7.37 ppm for 1H and 128.3 ppm for 13C).
1H NMR (400.1 MHz, CD3CN)
d
8.06 (td, J1¼7.6 Hz, J2¼0.7 Hz, 1H),
8.15e8.22 (m, 2H), 8.18 (d, J¼9.0 Hz, 1H), 8.32 (dd, J1¼8.0 Hz,
J2¼1.5 Hz, 1H), 8.49e8.57 (m, 2H), 8.58 (d, J¼9.2 Hz, 1H), 8.83 (d,
J¼9.0 Hz, 1H), 9.96 (d, J¼7.0 Hz, 1H); 1H NMR (500 MHz, DMSO-d6)
Reduction potentials E0 were based on energies calculated for
cations 1, radicals 1ꢃ, and anions 1L at the B3LYP/6-311þG(2d,p)//
B3LYP/6-311G(2d,p) level of theory with tight convergence criteria
in MeCN dielectric medium using the IPCM model37 (keywords:
SCRF¼IPCM) and epsilon value of 36.64. The free energies calcu-
lated with thermodynamic corrections derived from B3LYP/6-311G
d
8.08 (t, J¼7.5 Hz, 1H), 8.21 (td, J1¼8.0 Hz, J2¼1.6 Hz, 1H), 8.28 (td,
J1¼7.0 Hz, J2¼1.8 Hz, 1H), 8.39 (d, J¼9.0 Hz, 1H), 8.40 (dd, J1¼7.9 Hz,
J2¼1.5 Hz, 1H), 8.66 (td, J1¼7.7 Hz, J2¼0.9 Hz, 1H), 8.71 (dd,
J1¼8.3 Hz, J2¼1.6 Hz, 1H), 8.74 (d, J¼9.0 Hz, 1H), 9.16 (d, J¼8.9 Hz,
1H), 10.39 (d, J¼7.0 Hz, 1H); 13C NMR (75.5 MHz, DMSO-d6, DEPT
(2d,p) calculations were converted to volts using the expression DG
(sol)¼ꢀFE0 where F is the Faraday constant, and correcting by
4.189 V for the difference between the absolute potential for stan-
dard hydrogen electrode (SHE, 4.43 V)38 and SCE electrode relative
to SHE (0.241 V).
90) d 118.0, 122.9, 124.3, 126.8 (q), 128.4, 130.4, 130.5, 132.8, 134.2
(q),134.6, 136.6, 140.5, 143.3 (q); UV (CH3CN), lmax (log
227 (4.24), 253 (4.47), 277 (3.99), 299 (3.62), 332 (3.72), 347 (4.07),
363 (4.18) nm [lit.3 (EtOH) lmax (log
) 228 (3.85), 255 (4.07), 348
e) 222 (4.13),
e
Mechanistic studies were conducted at the B3LYP/6-311G(2d,p)
level of theory with tight convergence criteria. Global conforma-
tional minima for each trans and cis isomers, 3-E and 3-Z, were
established by comparison of SCF energies for four conformers at
fully optimized geometries without symmetry constraints. Transi-
tion state structures for cisetrans isomerization and formation of 6
were located using the STQN method39 requested with the QST2
keyword and default convergence criteria. Final energies for each
optimized structure were calculated with the B3LYP/6-311þG
(2d,p)//B3LYP/6-311G(2d,p) method and inclusion of solvation
effects requested with the SCRF¼IPCM keyword (epsilon¼36.64).
(3.68), 365 (3.80) nm]. Anal. Calcd for C13H10ClNO4: C, 55.83; H,
3.60; N, 5.01. Found: C, 55.83; H, 3.47; N, 4.99.
5.3.2. Pyrido[2,1-c][1,2,4]benzotriazin-11-ium
p-toluenosulfonate
(1b[OTs]). 2-(2-Fluorophenylazo)pyridine (3b, 201 mg, 1.0 mmol)
and calcium p-toluenosulfonate (190 mg, 1.0 mmol) were dissolved
in a MeCN/H2O mixture (9:1, 30 mL). The resulted solution was
irradiated with a halogen lamp and refluxed during 1e1.5 h (until
TLC control showed reaction was completed). Then, solvents were
evaporated. Residue was dried in a desiccator over P2O5 (12 h). The
solid was suspended in CH2Cl2 and filtered. Subsequently, it was

A. Jankowiak et al. / Tetrahedron 67 (2011) 3317e3327
3325
dissolved in hot MeCN, solution filtered, and the solvent was
evaporated to give a yellow-green solid. The product was recrys-
tallized (toluene/MeCN) to give 250 mg (85% yield) of 1b[OTs] as
(75.5 MHz, DMSO-d6, DEPT 90)
d
117.4, 121.9 (q), 124.6, 129.0, 130.8,
131.5, 134.1, 135.9 (q), 137.6, 144.2, 147.6 (q), 162.5; UV (CH3CN), lmax
(log ) 222 (4.01), 250 (4.54), 266 sh (4.23), 306 (3.66), 319 (3.77),
e
green-black crystals: mp 244e246 ꢁC; 1H NMR (CD3CN)
d
2.31 (s,
333.5 (3.98), 349 (4.04) nm. Anal. Calcd for C12H9ClN2O4: C, 51.35;
H, 3.23; N, 9.98. Found: C, 51.44; H, 3.34; N, 9.98.
3H), 7.12 (d, J ¼ 7.9 Hz, 2H), 7.56 (d, J ¼ 8.0 Hz, 2H), 8.37 (t, J ¼ 8.0 Hz,
1H), 8.51 (t, J ¼ 8.6 Hz,1H), 8.59 (t, J ¼ 7.0 Hz,1H), 8.92 (d, J ¼ 8.7 Hz,
1H), 9.08-9.03 (m, 2H), 9.27 (d, J ¼ 8.4 Hz, 1H), 10.08 (d, J ¼ 6.6 Hz,
5.3.6. trans-2-[2-(Chlorophenyl)vinyl]pyridine (2a)1. A mixture of
2-picoline (3.0 g, 32.30 mmol) and 2-chlorobenzaldehyde (4.74 g,
33.70 mmol) was refluxed in acetic anhydride (6 mL) for 20 h under
Ar atmosphere. Acetic anhydride was removed under vacuum,
a mixture of cyclohexane and EtOAc (1:1, 20 mL) was added, the
resulting solution was washed with satd NaHCO3, dried (Na2SO4),
and the solvents were evaporated. The crude product was purified
using a silica gel plug (CH2Cl2) and subsequent recrystallization
from hexane to give 4.73 g (69% yield) of 2a as pale yellow crystals:
1H); 13C NMR (75.5 MHz, DMSO-d6, DEPT 90)
d 20.8 (CH3), 116.9,
123.2 (q), 125.4, 128.0, 129.2, 130.4, 132.2, 137.6 (q), 138.8, 141.4 (q),
142.4 (q), 145.6 (q), 146.0. Anal. Calcd for C18H15N3O3S: C, 61.18; H,
4.28; N, 11.89. Found: C, 60.73; H, 4.12; N, 11.96.
5.3.3. Pyrido[2,1-c][1,2,4]benzotriazin-11-ium perchlorate (1b[ClO4])8.
Triazinium p-toluenosulfonate 1b[TsO] (100 mg, 0.28 mmol) was
dissolved in water (2 mL) and excess of satd solution of NaClO4 was
added. The resulting precipitation was filtrated to give 48 mg (61%
yield) of brown-red crystals: mp >295 ꢁC; 1H NMR (400 MHz, CD3CN)
mp 78e79 ꢁC (lit.1 mp 75e76.6 ꢁC); 1H NMR (400 MHz, CDCl3)
d 7.17
(d, J¼16.1 Hz, 1H), 7.15e7.20 (m, 1H), 7.23 (td, J1¼7.7 Hz, J2¼1.8 Hz,
1H), 7.29 (td, J1¼7.4 Hz, J2¼1.4 Hz,1H), 7.41 (dd, J1¼7.8 Hz, J2¼1.4 Hz,
1H), 7.47 (d, J¼7.9 Hz, 1H), 7.68 (td, J1¼7.7 Hz, J2¼1.8 Hz, 1H), 7.73
(dd, J1¼7.7 Hz, J2¼1.8 Hz, 1H), 7.98 (d, J¼16.2 Hz, 1H), 8.63 (dd,
d
8.41 (d, J ¼ 7.7 Hz,1H), 8.53 (td, J1 ¼8.7 Hz, J2 ¼1.4 Hz,1H), 8.60 (t, J ¼
6.3 Hz,1H), 8.86 (d, J ¼ 8.7 Hz,1H), 9.08 (t, J¼ 7.6 Hz,1H), 9.30 (d, J ¼ 8.4
Hz, 1H), 9.97 (d, J ¼ 6.7 Hz, 1H); 1H NMR (500 MHz, DMSO-d6)
d 8.42
(td, J1 ¼8.0 Hz, J2 ¼ 0.6 Hz,1H), 8.60 (td, J1 ¼8.7 Hz, J2 ¼1.4 Hz,1H), 8.74
(td, J1 ¼7.6 Hz, J2 ¼1.5 Hz,1H), 9.13 (dd, J1 ¼8.1 Hz, J2 ¼1.3 Hz,1H), 9.18
(td,J1 ¼8.94, Hz, J2 ¼1.1 Hz,1H), 9.23 (d, J¼8.7Hz,1H), 9.46(dd,J1 ¼6.9
Hz, J2 ¼ 2.1 Hz, 1H), 10.50 (d, J ¼ 6.6 Hz, 1H); 13C NMR (75.5 MHz,
J1¼4.8 Hz, J2¼0.8 Hz, 1H); 1H NMR (500 MHz, CD3CN)
d 7.24e7.27
(m, 1H), 7.28 (d, J¼16.0 Hz, 1H), 7.31 (td, J1¼7.6 Hz, J2¼1.7 Hz, 1H),
7.37 (td, J1¼7.9 Hz, J2¼0.7 Hz, 1H), 7.470 (d, J¼8.9 Hz, 1H), 7.474 (d,
J¼8.9 Hz, 1H), 7.76 (td, J1¼7.7 Hz, J2¼1.8 Hz, 1H), 7.86 (dd, J1¼7.8 Hz,
J2¼1.5 Hz, 1H), 8.07 (d, J¼15.7 Hz, 1H), 8.61 (d, J¼4.2 Hz, 1H).
DMSO-d6, DEPT 90) d 116.8, 123.1 (q), 129.2, 130.5, 132.3, 133.0, 133.3,
138.9, 141.4 (q), 142.4 (q), 146.0; UV (CH3CN), lmax (log
e) 259 (4.32),
347 (3.87), 359 (3.88) [lit.8 (CH3CN), lmax (log
e) 258 (4.37), 348 (3.90),
5.3.7. 2-(2-Chlorophenylazo)pyridine (2b)12. To the solution of
2-aminopyridyne, (67 mg, 0.7 mmol) in toluene (0.1 mL), a solution
of NaOH (50% in water, 0.7 mL) followed by 2-chloroni-
trosobenzene (102 mg, 0.7 mmol) was added. The mixture was
vigorously stirred with mechanic stirrer for 25 min at 50 ꢁC. After
cooling, water was added and mixture was extracted with CH2Cl2,
combined extracts were dried (Na2SO4), and the solvents were
evaporated. The residue was purified using a silica gel plug (hex-
ane/CH2Cl2, 5:1) followed by recrystallization from hexane to give
69 mg (61% of yield) of 2b as red crystals: mp 52e54 ꢁC; (lit.12 mp
359 (3.91), 375 sh (3.73)]. Anal. Calcd for C11H8ClN3O4: C, 46.91; H,
2.86; N, 14.92. Found: C, 47.04; H, 2.78; N, 14.70.
5.3.4. Pyrido[1,2-a]quinoxalin-11-ium perchlorate (1c[ClO4]). Using
a procedure similar to that described for 1d[ClO4], salt 1c[ClO4] was
prepared in 12% yield from 3c as red crystals: mp 246e248 ꢁC; 1H
NMR (400 MHz, CD3CN)
d 8.15e8.23 (m, 2H), 8.43e8.47 (m, 1H),
8.52 (td, J1¼7.0 Hz, J2¼1.6 Hz, 1H), 8.78e8.83 (m, 2H), 8.87 (td,
J1¼7.7 Hz, J2¼1.0 Hz, 1H), 9.66 (s, 1H), 10.05 (d, J¼6.8 Hz, 1H); 1H
NMR (500 MHz, DMSO-d6)
d
8.18 (td, J1¼7.5 Hz, J2¼1.1 Hz, 1H), 8.22
54e55 ꢁC); 1H NMR (CDCl3)
d 8.77 (m, 1H), 7.94e7.84 (m, 3H), 7.58
(d, J¼8.0 Hz, 1H), 7.49e7.42 (m, 2H), 7.36 (t, J¼7.9 Hz, 1H).
(td, J1¼7.2 Hz, J2¼1.6 Hz, 1H), 8.44 (dd, J1¼7.8 Hz, J2¼1.7 Hz, 1H),
8.64 (td, J1¼6.9 Hz, J2¼2.1 Hz, 1H), 8.99 (td, J1¼8.0 Hz, J2¼0.7 Hz,
1H), 9.01 (dd, J1¼8.0 Hz, J2¼2.0 Hz, 1H), 9.15 (d, J¼8.3 Hz, 1H), 9.88
(s, 1H), 10.54 (d, J¼6.7 Hz, 1H); 13C NMR (75.5 MHz, DMSO-d6, DEPT
5.3.8. trans-2-[2-(Fluorophenyl)vinyl]pyridine (3a)11. Prepared sim-
ilarly to 2a, 3a was obtained in 34% yield as brown-yellow crystals:
90)
d
117.7, 126.7 (q), 128.7, 129.1, 131.2, 132.4, 132.5, 134.4 (q), 134.9,
mp 74e76 ꢁC (lit.11 mp 71e72 ꢁC); 1H NMR (400 MHz, CDCl3)
d 7.09
139.2 (q), 143.7, 150.6; UV (CH3CN), lmax (log
e
) 226 (4.08), 242
(ddd, J1¼10.8 Hz, J2¼8.2 Hz, J3¼1.1 Hz, 1H), 7.16 (br d, J¼7.5 Hz, 1H),
7.15e7.19 (m, 1H), 7.23e7.28 (m, 1H), 7.28 (d, J1¼16.4 Hz, 1H), 7.43
(br d, J¼7.9 Hz, 1H), 7.64 (td, J1¼7.7 Hz, J2¼1.7 Hz, 1H), 7.67 (td,
J1¼7.8 Hz, J2¼1.8 Hz,1H), 7.76 (d, J¼16.3 Hz, 1H), 8.62 (dd, J1¼4.8 Hz,
(4.14), 256 (4.25), 317 (3.50), 331 (3.72), 346.5 (3.99), 363.5 (4.07)
nm. Anal. Calcd for C12H9ClN2O4: C, 51.35; H, 3.23; N, 9.98. Found: C,
51.18; H, 3.09; N, 9.97.
J2¼0.8 Hz, 1H); 1H NMR (500 MHz, CD3CN)
d
7.17 (dd, J1¼10.8 Hz,
5.3.5. Pyrido[1,2-a]quinazolin-11-ium perchlorate (1d[ClO4]). A
mixture of imine 3d (3.80 g, 19.0 mmol) and Ca(OTs)2 (3.20 g,
8.4 mmol) in dry PhCN (60 mL) was heated at 160 ꢁC for 3 h. The
solvent was removed at reduced pressure and the residue was
dissolved in water. An insoluble black solid was filtered, the aque-
ous filtrate was heated with activated carbon, filtered, concentrated
in vacuum and excess of satd solution of NaClO4 was added. The
resulting precipitation was filtered and dried to give 1.23 g (18%
yield) of 1d[ClO4] as red crystals. An analytically pure sample was
obtained by recrystallization of the salt from EtOH and then from
J2¼8.6 Hz, 1H), 7.21e7.26 (m, 1H), 7.24 (d, J¼7.6 Hz, 1H), 7.32e7.38
(m, 1H), 7.34 (d, J¼16.2 Hz, 1H), 7.48 (d, J¼7.8 Hz, 1H), 7.75 (dd,
J1¼7.7 Hz, J2¼1.7 Hz, 1H), 7.78 (dd, J1¼7.7 Hz, J2¼1.5 Hz, 1H), 7.83 (d,
J¼16.2 Hz, 1H), 8.60 (d, J¼4.4 Hz, 1H).
5.3.9. 2-(2-Fluorophenylazo)pyridine (3b). Prepared similarly to that
for 2b, 3b was obtained in 71% yield as red crystals: mp 52e54 ꢁC;
1H NMR (CDCl3)
d
8.76 (dm, J¼5.1 Hz, 1H), 7.94e7.83 (m, 3H),
7.56e7.49 (m,1H), 7.43 (t, J¼6.2 Hz, 1H), 7.33e7.22 (m, 2H); 13C NMR
(75.5 MHz, CDCl3, DEPT 90)
d
114.2,117.1 (d, J¼20 Hz), 117.8,124.3 (d,
MeCN: mp 244e246 ꢁC; 1H NMR (400 MHz, CD3CN)
d
8.21 (td,
J¼4 Hz), 125.5, 133.3 (d, J¼9 Hz), 138.3, 140.2 (q), 149.5, 160.7 (q,
J¼260 Hz), 163.0 (q). Anal. Calcd for C11H8FN3: C, 65.67; H, 4.01; N,
20.88. Found: C, 65.88; H, 4.07; N, 20.62.
J1¼7.6 Hz, J2¼0.5 Hz, 1H), 8.25 (td, J1¼7.1 Hz, J2¼1.5 Hz, 1H), 8.45
(ddd, J1¼8.9 Hz, J2¼7.3 Hz, J3¼1.5 Hz, 1H), 8.53 (dd, J1¼7.9 Hz,
J2¼1.4 Hz, 1H), 8.59 (dd, J1¼8.4 Hz, J2¼1.2 Hz, 1H), 8.74 (ddd,
J1¼8.5 Hz, J2¼7.3 Hz, J3¼1.3 Hz, 1H), 8.83 (d, J¼8.9 Hz, 1H), 9.80 (s,
5.3.10. 2-Fluoro-N-(pyridin-2-ylmethylene)aniline (3c). Using anal-
ogous procedure described for 3d, imine 3c was prepared in 55%
yield as pale yellow crystals: mp 50e51 ꢁC; 1H NMR (400 MHz,
1H), 9.86 (d, J¼6.9 Hz, 1H); 1H NMR (500 MHz, DMSO-d6)
d 8.23 (t,
J¼7.5 Hz, 1H), 8.36 (td, J1¼7.0 Hz, J2¼1.2 Hz, 1H), 8.49 (td, J1¼8.0 Hz,
J2¼1.1 Hz, 1H), 8.63 (d, J¼7.9 Hz, 1H), 8.68 (d, J¼8.4 Hz, 1H), 9.17 (d,
J¼8.9 Hz, 1H), 10.00 (s, 1H), 10.32 (d, J¼6.8 Hz, 1H); 13C NMR
CDCl3)
d
7.13e7.25 (m, 4H), 7.39 (ddd, J1¼7.5 Hz, J2¼4.8 Hz, J3¼1.2 Hz,
1H), 7.83 (td, J1¼7.7 Hz, J2¼1.5 Hz, 1H), 8.27 (d, J¼7.9 Hz,1H), 8.67 (s,

3326
A. Jankowiak et al. / Tetrahedron 67 (2011) 3317e3327
1H), 8.72 (dq, J1¼4.8 Hz, J2¼0.9 Hz, 1H); 13C NMR (75 MHz, CDCl3)
References and notes
d
116.3 (d, 2JCF¼20 Hz), 121.7, 121.9, 124.5 (d, 4JCF¼4 Hz), 125.4, 127.5
(d, 3JCF¼8 Hz), 136.6, 139.0 (q) (d, 3JCF¼10 Hz), 149.6, 154.3, 155.3 (d,
1JCF¼250 Hz), 163.0. Anal. Calcd for C12H9FN2: C, 71.99; H, 4.53; N,
13.99. Found: C, 72.23; H, 4.48; N, 14.06.
1. Fozard, A.; Bradsher, C. K. J. Org. Chem. 1966, 31, 2346e2349.
2. Fozard, A.; Davies, L. S.; Bradsher, C. K.; Gross, P. M. J. Chem. Soc. C 1971,
3650e3652.
3. Arai, S.; Arai, H.; Tabuchi, K.; Yamagishi, T.; Hida, M. J. Heterocycl. Chem. 1992,
29, 215e220.
4. Arai, S.; Arai, H.; Hida, M.; Yamagishi, T. Heterocycles 1994, 38, 2449e2454.
5. Norez, C.; Bilan, F.; Kitzis, A.; Mattey, Y.; Becq, F. J. Pharmcol. Exp. Therap.
2008, 325, 89e99; Roomans, G. M. Am. J. Respir. Med. 2003, 2, 413e431;
5.3.11. 2-(2-Fluorobenzylidenamino)pyridine (3d). A mixture of
2-fluorobenzaldehyde (136 mg, 1.1 mmol), 2-aminopyridine
(94 mg, 1 mmol), and TsOH (cat.) in benzene (5 mL) was refluxed
for 4 h, the cooled solution was washed with satd NaHCO3, dried
(NaSO4), and solvents were evaporated. The crude mixture was
recrystallized (hexane) to give 135 mg (68% yield) of 3d as pale
yellow crystals: mp 120e122 ꢁC; 1H NMR (400 MHz, CDCl3)
ꢀ
Prost, A.-L.; Derand, R.; Gros, L.; Becq, F.; Vivaudou, M. Biochem. Pharmcol.
2003, 66, 425e430; Galietta, L. J. V.; Springsteel, M. F.; Eda, M.; Niedzinski,
E. J.; By, K.; Haddadin, M. J.; Kurth, M. J.; Nantz, M. H.; Verkman, A. S. J. Biol.
Chem. 2001, 276, 19723e19728.
€
6. Wu, D.; Pisula, W.; Enkelmann, V.; Feng, X.; Mullen, K. J. Am. Chem. Soc. 2009,
131, 9620e9621.
7. Garza-Ortiz, A.; den Dulk, H.; Brouwer, J.; Kooijman, H.; Spek, A. L.; Reedijk, J. J.
Inorg. Biochem. 2007, 101, 1922e1930.
8. Sinan, M.; Panda, M.; Ghosh, A.; Dhara, K.; Fanwick, P. E.; Chattopadhyay, D. J.;
Goswami, S. J. Am. Chem. Soc. 2008, 130, 5185e5193.
d
7.12e7.17 (m, 1H), 7.19 (ddd, J1¼7.3 Hz, J2¼4.9 Hz, J3¼0.9 Hz, 1H),
7.25 (t, J¼7.6 Hz, 1H), 7.34 (d, J¼8.0 Hz, 1H), 7.45e7.52 (m, 1H), 7.76
(td, J1¼7.7 Hz, J2¼1.9 Hz, 1H), 8.25 (td, J1¼7.5 Hz, J2¼1.7 Hz, 1H), 8.51
9. Koner, R. R.; Ray, M. Inorg. Chem. 2008, 47, 9122e9124.
10. Manka, J. T.; McKenzie, V. C.; Kaszynski, P. J. Org. Chem. 2004, 69, 1967e1971.
(dd, J1¼4.8 Hz, J2¼1.1 Hz, 1H), 9.45 (s, 1H); 13C NMR (75.5 MHz,
2
€
11. Schubert, R.; Ramana, D. V.; Grutzmacher, H.-F. Chem. Ber. 1980, 113, 3758e3774.
CDCl3, DEPT-90)
d
115.9 (d, JCF¼21 Hz), 119.7, 122.0, 123.6(q) (d,
2JCF¼9 Hz), 124.3 (d, JCF¼4 Hz), 128.1 (d, JCF¼2 Hz), 133.5 (d,
3
4
12. Campbell, N.; Henderson, A. W.; Taylor, D. J. Chem. Soc. 1953, 1281e1285.
13. Darchen, A.; Peltier, D. Bull. Soc. Chim. Fr. 1974, 673e676.
3JCF¼9 Hz), 138.0, 148.9, 156.2 (d, JCF¼5 Hz), 161.0(q), 163.3 (d,
3
14. Crystal data for 1a[ClO4] (CCDC no. 792984): C13H10ClNO4 triclinic, Pꢀ1, a¼9.
1JCF¼255 Hz). Anal. Calcd for C12H9FN2: C, 71.99; H, 4.53; N, 13.99.
6517(12) A, b¼11.0470(13) A, c¼12.2373(15) A,
a
¼67.615(1)ꢁ,
b
¼78.845(2)ꢁ,
2
ꢁ
ꢁ
3
ꢁ
ꢁ
ꢁ
ꢁ
g
¼87.559(2) ; V¼1183.0(2) A , Z¼4, T¼173(2) K,
l
¼0.71073 A, R(F )¼0.0373 or
Found: C, 71.86; H, 4.67; N, 13.99.
Rw(F2)¼0.0971 (for 4200 reflections with I>2
s(I)). Crystal data for 1d[ClO4]
ꢁ
ꢁ
3
ꢁ
(CCDC no. 792983): C12H9ClN2O4 triclinic, Pꢀ1, a¼5.9525(6) A, b¼8.3141(9) A,
5.3.12. 2-Chloronitrosobenzene (4)12. Prepared similarly to 5, 4 was
obtained in 60% yield as yellow crystals: mp 64.5e65.5 ꢁC (lit.12 mp
ꢁ
¼73.487(1)ꢁ,
b
¼83.814(1)ꢁ,
g¼83.456(1) ; V¼576.07(10) A ,
ꢁ
c¼12.2591(13) A,
a
2
ꢁ
Z¼2, T¼173(2) K,
l
¼0.71073 A, R(F )¼0.0414 or Rw(F2)¼0.1096 (for 2287 re-
flections with I>2s(I)). For details see Supplementary data.
65.5e66.5 ꢁC); 1H NMR (300 MHz, CDCl3)
7.49e7.42 (m, 2H), 7.58 (d, J¼8.0 Hz, 1H), 7.94e7.84 (m, 3H), 8.77
(dd, J1¼1.6 Hz, J2¼4.8 Hz, 1H). Anal. Calcd for C6H4ClNO: C, 50.91; H,
2.85; N, 9.89. Found: C, 50.93; H, 2.78; N, 9.89.
d
7.36 (t, J¼7.9 Hz, 1H),
15. Smiglak, M.; Hines, C. C.; Wilson, T. B.; Singh, S.; Vincek, A. S.; Kirichenko, K.;
Katritzky, A. R.; Rogers, R. D. Chem.dEur. J. 2010, 16, 1572e1584.
16. For details see Supplementary data.
17. Wang, L.; Yi, C.; Zou, H.; Xu, J.; Xu, W. J. Phys. Org. Chem. 2009, 22, 888e896.
18. Crecca, C. R.; Roitberg, A. E. J. Phys. Chem. A 2006, 110, 8188e8203.
19. Gaenko, A. V.; Devarajan, A.; Gagliardi, L.; Lindh, R.; Orlandi, G. Theor. Chem. Acc.
2007, 118, 271e279.
20. Ammal, S. C.; Yamataka, H. Eur. J. Org. Chem. 2006, 4327e4330.
21. Geibel, K.; Staudinger, B.; Grellmann, K. H.; Wendt, H. Ber. Bunsen-Ges. Phys.
Chem. 1972, 76, 1246e1251.
5.3.13. 2-Fluoronitrosobenzene (5)13. To a solution of 2-fluoroani-
line (0.90 mL, 9.30 mmol) in CH2Cl2 (20 mL), a water (80 mL)
solution of OxoneÒ (11.0 g, 17.90 mmol) was added, and the re-
action stirred at rt under N2 atmosphere, with exclusion of light, for
24 h. The organic layer was separated and the aqueous layer was
extracted with CH2Cl2. The combined organic layers were washed
with 5% of HCl followed by satd NaHCO3 and then water, dried
(NaSO4), and the solvents were evaporated. The residue was passed
through a silica gel plug (hexane/CH2Cl2, 9:1) and crystallized
(MeOH/H2O) to give 460 mg (41% yield) of 5 as pale yellow crystals:
22. Maeda, K.; Fischer, E. Helv. Chim. Acta 1983, 66, 1961e1965.
23. Brown, E. V.; Granneman, G. R. J. Am. Chem. Soc. 1975, 97, 621e627.
24. Kistiakowsky, G. B.; Smith, W. R. J. Am. Chem. Soc. 1934, 56, 638e642.
25. All reported spectra were obtained in anhydrous DMSO-d6 taken from freshly
opened ampule. Same spectra obtained in regular DMSO-d6 with a relatively
large water content show downfield shift up to 0.15 ppm for some protons.
26. Kosower, E. M. Top. Curr. Chem. 1983, 112, 117e162.
27. Becke, A. D. J. Chem. Phys. 1993, 98, 5648e5652.
28. Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785e789.
29. Møller, C.; Plesset, M. S. Physiol. Rev. 1934, 46, 618e622; Head-Gordon, M.;
mp 68e70 ꢁC (lit.13 mp 66e68 ꢁC); 1H NMR (300 MHz, CD3Cl)
d 6.49
Pople, J. A.; Frisch, M. J. Chem. Phys. Lett. 1988, 153, 503e506.
(ddd, J1¼10.5 Hz, J2¼8.5 Hz, J3¼1.8 Hz, 1H), 7.14 (t, J¼7.0 Hz, 1H),
7.51 (ddd, J1¼8.0 Hz, J2¼7.0 Hz, J3¼1.1 Hz, 1H), 7.76e7.68 (m, 1H).
Anal. Calcd for C6H4FNO: C, 57.61; H, 3.22; N, 10.94. Found: C, 56.95;
H, 3.34; N, 10.94.
30. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J.
C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi,
M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara,
M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.;
Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J.
B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.;
Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.;
Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.;
Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck,
A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.;
Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.;
Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.;
Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.;
Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, Revision D.01; Gauss-
ian,: Wallingford CT, 2004.
31. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. K.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Na-
katsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng,
G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.;
Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery,
J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K.
N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.;
Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.;
Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.;
Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.;
Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dan-
nenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.; Foresman, J. B.; Ortiz, J. V.;
Cioslowski, J.; Fox, D. J. Gaussian 09, Revision A.02; Gaussian,: Wallingford CT,
2009.
Acknowledgements
Financial support for this work was received from the Petroleum
Research Funds (PRF-47243-AC4) and NSF (OISE 0532040).
Supplementary data
These data include 2-D NMR spectra for 1[ClO4], tabularized
theoretical interatomic distances for 1, distribution of spin density
for 1ꢃ, crystal data for 1a[ClO4] and 1d[ClO4], archive of calculated
equilibrium geometries for 1 and 1ꢃ, 1ꢀ, 2-E, 2-Z, 1-TS, 1-Cl, 3-E, 3-Z,
3-TS, 6, and 6-TS.
CCDC 792984 and 792983 contains the supplementary crystal-
lographic data for 1a[ClO4] and 1d[ClO4]. These data can be obtained
freeof charge via http://www.ccdc.cam.ac.uk/conts/retrieving.html,
or from the Cambridge Crystallographic Data Centre,12 Union Road,
Cambridge CB2 1EZ, UK; fax: (þ44) 1223 336 033; or e-mail:
deposit@ccdc.cam.ac.uk. Supplementary data associated with this
article can be found in the online version, at doi:10.1016/
j.tet.2011.03.023.
32. Scott, A. P.; Radom, L. J. Phys. Chem. 1996, 100, 16502e16513.

A. Jankowiak et al. / Tetrahedron 67 (2011) 3317e3327
3327
33. Stratmann, R. E.; Scuseria, G. E.; Frisch, M. J. J. Chem. Phys. 1998, 109,
8218e8224.
37. Foresman, J. B.; Keith, T. A.; Wiberg, K. B.; Snoonian, J.; Frisch, M. J. J. Phys. Chem.
1996, 100, 16098e16104.
34. Cossi, M.; Scalmani, G.; Rega, N.; Barone, V. J. Chem. Phys. 2002, 117, 43e54 and
references therein.
35. Wolinski, K.; Hinton, J. F.; Pulay, P. J. Am. Chem. Soc. 1990, 112, 8251e8260 and
references therein.
38. Reiss, H.; Heller, A. J. Phys. Chem. 1985, 89, 4207e4213.
39. Peng, C.; Ayala, P. Y.; Schlegel, H. B.; Frisch, M. J. J. Comput. Chem. 1996, 17,
49e56.
40. Noviandri, I.; Brown, K. N.; Fleming, D. S.; Gulyas, P. T.; Lay, P. A.; Masters, A. F.;
Phillips, L. J. Phys. Chem. B 1999, 103, 6713e6722.
36. Gottlieb, H. E.; Kotlyar, V.; Nudelman, A. J. Org. Chem. 1997, 62, 7512e7515.