Substituent-dependent negative hyperconjugation in 2-aryl-1,3-N,N-heterocycles. Fine-tuned anomeric effect?

Article abstract of DOI:10.1021/jo034417+

The epimerization reactions of conformationally inflexible 2-aryl-1,3-N,N-heterocycles were used as model systems to study the role of the nitrogen lone pair-C2 associated antibonding orbital hyperconjugative interactions in the experimentally observed substituent-dependent generalized anomeric effect. The measured reaction free enthalpies were found to correlate well with the sum of the hyperconjugative stabilization energies of all the vicinal donor - acceptor orbital overlaps around C2, obtained from ab initio NBO analysis, and both quantities correlated linearly with the Hammett-Brown substituent constant. The individual stereoelectronic interactions (n_N-σ*_C2-N, n_N-σ*_C2-Ar, n_N-σ*_C2-H) were also observed to exhibit a substituent dependence, despite their distance from the 2-aryl substituent and their nonperiplanar arrangement. The higher the electron-withdrawing effect of the 2-aryl substituent, the larger was the stabilization for n_N-σ*_C2-Ar, while the overlaps n_N-σ*_C2-N and n_N-σ*_C2-H changed in the opposite sense. The different polarization of the acceptor σ* orbitals, caused by the 2-aryl substituent, accounted for the observed propagation of the substituent effect. These results promote a detailed explanation of the useful tautomeric behavior of the 2-aryl-1,3-X,N-heterocycles, and reveal the nature of the connection between the anomeric effect and the Hammett-type linear free energy relationship.

Full text of DOI:10.1021/jo034417+

Su bstitu en t-Dep en d en t Nega tive Hyp er con ju ga tion in
2-Ar yl-1,3-N,N-h eter ocycles. F in e-Tu n ed An om er ic Effect?
Anaszta´zia Hete´nyi, Tama´s A. Martinek, La´szlo´ La´za´r, Zita Zala´n, and Ferenc Fu¨lo¨p*
Institute of Pharmaceutical Chemistry, University of Szeged, H-6701 Szeged, POB 121, Hungary
fulop@pharma.szote.u-szeged.hu
Received April 1, 2003
The epimerization reactions of conformationally inflexible 2-aryl-1,3-N,N-heterocycles were used
as model systems to study the role of the nitrogen lone pair-C2 associated antibonding orbital
hyperconjugative interactions in the experimentally observed substituent-dependent generalized
anomeric effect. The measured reaction free enthalpies were found to correlate well with the sum
of the hyperconjugative stabilization energies of all the vicinal donor-acceptor orbital overlaps
around C2, obtained from ab initio NBO analysis, and both quantities correlated linearly with the
Hammett-Brown substituent constant. The individual stereoelectronic interactions (nN-σ*C2-N
,
nN-σ*C2-Ar, nN-σ*C2-H) were also observed to exhibit a substituent dependence, despite their
distance from the 2-aryl substituent and their nonperiplanar arrangement. The higher the electron-
withdrawing effect of the 2-aryl substituent, the larger was the stabilization for nN-σ*C2-Ar, while
the overlaps nN-σ*C2-N and nN-σ*C2-H changed in the opposite sense. The different polarization of
the acceptor σ* orbitals, caused by the 2-aryl substituent, accounted for the observed propagation
of the substituent effect. These results promote a detailed explanation of the useful tautomeric
behavior of the 2-aryl-1,3-X,N-heterocycles, and reveal the nature of the connection between the
anomeric effect and the Hammett-type linear free energy relationship.
In tr od u ction
studies involving the use of various functional groups as
Y and elements with intermediate EN as A.4,6,7 Despite
the difficulties in the detection of the low energy differ-
ences in systems containing carbon as Y (e.g., -COR,
-COOR, or -Ar),8-10 it has been revealed that the
stability difference between the C2 epimers in the
epimerization equilibrium of 2-aryl-trans-1,3-X,S-decalins
is affected by the electronic properties of the aromatic
substituent.11,12 In regard to the local geometry, observa-
tions of AE have been reported in conformations other
than antiperiplanar.13
In this work, we intend to extend the application of
the concept of AE. Using the epimerization equilibrium
of 2-aryl-1,3-N,N-heterocycles as model system, we show
that substituent-dependent stereoelectronic stabilization
can occur even in a nonperiplanar arrangement and it
can be characterized by the Hammett-Brown substitu-
ent constants by making use of linear free energy
The generalized anomeric effect (AE) is recognized in
an Lp-X-A-Y moiety as a preference for an anti-
periplanar arrangement of the lone pair (Lp) and the
A-Y bond, where X is a heteroatom, A is an element with
intermediate electronegativity (EN), and Y is a group
with higher EN.1,2 Despite the dispute continuing up to
recent years concerning the exact origin of this well-
established conformational phenomenon,3,4 it is generally
accepted that the stability difference observed between
two specific geometries of the same molecular moiety
stems from two-electron/two-orbital hyperconjugative
interactions resulting in an excess stabilization energy.5
There are four major factors that influence the excess of
the electron delocalization energy: (i) the EN difference
along the A-Y bond, (ii) the acceptor orbital energy level,
(iii) the nature of the atom X, and (iv) the Lp-X-A-Y
dihedral angles observed in the compared structures.
Handling the stability difference as a function of the
acceptor ability of the A-Y bond and that of the Lp-X-
A-Y dihedral angle allows a further generalization of the
concept. This approach has resulted in a number of
(6) Mikolajczyk, M.; Garaczyk, P.; Wieczorek, M. W.; Bujacz, G.
Angew. Chem., Int. Ed. Engl. 1991, 30, 578-580.
(7) (a) Reed, A. E.; Schleyer, P. v. R. J . Am. Chem. Soc. 1987, 109,
7362-7373. (b) Alabugin, I. V.; Zeidan, T. A. J . Am. Chem. Soc. 2002,
124, 3475-3185.
(8) Anderson, C. B.; Sepp, D. T. J . Org. Chem. 1968, 33, 3272-3276.
(9) J uaristi, E.; Tapia, J .; Me´ndez, R. Tetrahedron 1986, 42, 1253-
1264.
* To whom correspondence should be addressed. Tel: +36-62-
545564. Fax: +36-62-545705.
(1) Kirby, A. J . The Anomeric Effect and Related Stereoelectronic
Effects at Oxygen; Springer-Verlag: Berlin, 1983.
(2) J uaristi, E.; Cuevas, G. Tetrahedron 1992, 48, 5019-5087.
(3) Salzner, U. J . Org. Chem. 1995, 60, 986-995.
(4) Carballeira, L.; Pe´rez-J uste, I. J . Phys. Chem. 2000, 104, 9362-
9369.
(10) J uaristi, E.; Gonza´lez, E. A.; Pinto, B. M.; J ohnston, B. D.;
Nagelkerke, R. J . Am. Chem. Soc. 1989, 111, 6745-6749.
(11) Tschierske, C.; Ko¨hler, H.; Zaschke, H.; Kleinpeter, E. Tetra-
hedron 1989, 45, 6987-6998.
(12) Ko¨hler, H.; Tschierske, C.; Zaschke, H.; Kleinpeter, E. Tetra-
hedron 1990, 46, 4241-4246.
(5) (a) Alabugin, I. V. J . Org. Chem. 2000, 65, 3910-3919. (b)
Schleyer, P. v. R.; Kos, A. J . Tetrahedron 1983, 39, 1141-1150.
(13) King, J . F.; Rathore, R.; Guo, Z.; Li, M.; Payne, N. C. J . Am.
Chem. Soc. 2000, 122, 10308-10324.
10.1021/jo034417+ CCC: $25.00 © 2003 American Chemical Society
Published on Web 06/06/2003
J . Org. Chem. 2003, 68, 5705-5712
5705

Hete´nyi et al.
SCHEME 1
SCHEME 2
relationship (LFER) analysis. The results of the ab initio
calculations performed, including the natural bond or-
bital (NBO)14-17 method, shed considerable light on the
mechanism of propagation of the influence of the 2-aryl
substituent on the ring stability of the 1,3-N,N-hetero-
cycles due to substituent-dependent hyperconjugative
interactions.
SCHEME 3
Resu lts a n d Discu ssion
Th e Mod el Com p ou n d s. The 1,3-X,N-heterocycles
are apt to participate in equilibrium processes involving
opening of the heterocyclic ring at C2. These processes
include ring-ring epimerization and ring-chain tautom-
erism.18 The early works of Kleinpeter et al.,11,12 our own
previous studies on 2-aryl-1,3-X,N-heterocycles,19 and the
recent work of Neuvonen et al.20 turned our attention to
the fact that not only the ring-chain equilibrium ratios
but also the ring-ring epimerization equilibrium can be
influenced by the 2-aryl substituent. The substituent
dependence of the ring stability is affected by the relative
configuration of C2, and thus, this stability difference
presumably caused by an AE can be separated experi-
mentally. We assumed that the observation of the sub-
stituent-dependent AE is related to the conformational
rigidity of the 1,3-N,N-heterocycles caused by the con-
densed ring system and the tertiary nitrogen substituted
by sterically demanding groups. Therefore, inflexible
model compounds 11-14 containing a tertiary nitrogen
in ring annelation were synthesized from the correspond-
ing 1,2,3,4-tetrahydroisoquinoline-1- or 3-methylamines.21
3-Aminomethyl-1,2,3,4-tetrahydroisoquinoline 3, neces-
sary for the linear imidazoisoquinoline model compounds,
was synthesized by reduction of the corresponding car-
boxamide 2 (Scheme 1) obtained from the easily available
amino ester 1.22
In contrast with the great number of imidazo[5,1-a]- and
imidazo[1,5-b]isoquinolines bearing oxo/thioxo substitu-
ents or sp2 carbons at positions 1 and/or 3,24,25 only one
recent publication is known that relates to the synthesis
of the saturated analogues with sp3 carbons in those
positions.26
The condensations of diamines 3 and 8 with equivalent
amounts of nine aromatic aldehydes in CDCl3 at 300 K
resulted in three-component tautomeric mixtures 9, 11,
12 and 10, 13, 14 (two C-3 epimeric imidazoisoquinolines
and the corresponding Schiff base) (Scheme 3), containing
angularly and linearly condensed imidazoisoquinolines
in epimerization equilibria.
1-Aminomethyl-6,7-dimethoxy-1,2,3,4-tetrahydroiso-
quinoline 8, the starting material for the synthesis of
angularly condensed imidazoisoquinoline model com-
pounds, was prepared in four steps, starting from N-
protected glycine 4 and homoveratrylamine (Scheme 2).
(14) Salzner, U.; Schleyer, P. R. J . Org. Chem. 1994, 59, 2138-2155.
(15) Corte´s, F.; Tenorio, J .; Collera, O.; Cuevas, G. J . Org. Chem.
2001, 66, 2918-2924.
(23) Compounds 1-3 and 7-14 are racemates. In the schemes, only
the enantiomers of 10-14 having the annelation hydrogen in the â
position are shown.
(16) Reed, E. A.; Weinstock, R. B.; Weinhold, F. J . Chem. Phys. 1985,
83, 735-746.
(24) (a) Liao, Z.; Kohn, H. J . Org. Chem. 1984, 49, 4745-4752. (b)
Kano, S.; Yuasa, Y.; Shibuya, S. Synth. Commun. 1985, 15, 883-889.
(c) J ones, R. C. F.; Smallridge, M. J .; Chapleo, C. B. J . Chem. Soc.,
Perkin Trans. 1 1990, 385-391. (d) Fu¨lo¨p, F.; Semega, EÄ .; Berna´th,
G.; Soha´r, P. Pharmazie 1992, 47, 168-171. (e) Collado, M. I.;
Sotomayor, N.; Villa, M.; Lete, E. Tetrahedron Lett. 1996, 37, 6193-
6196. (f) Osante, I.; Collado, M. I.; Lete, E.; Sotomayor, N. Eur. J . Org.
Chem. 2001, 1267-1277. (g) Funabashi, K.; Ratni, H.; Kanai, M.;
Shibasaki, M. J . Am. Chem. Soc. 2001, 123, 10784-10785.
(25) (a) Weir, E. C.; Niopas, I.; Smail, G. A. Chem. Chron. 1989, 18,
3-17. (b) Albanoudi, A.; Gabrieli, C.; Niopas, I. Sci. Pharm. 1990, 58,
259-262. (c) Niopas, I.; Smail, G. A. J . Chem. Soc., Perkin Trans. 1
1991, 113-117. (d) Wu, S.; J anusz, J . M.; Sheffer, J . B. Tetrahedron
Lett. 2000, 41, 1159-1163. (e) Charton, J .; Delarue, S.; Vendeville, S.;
Debreu-Fontaine, M.; Girault-Mizzi, S.; Sergheraert, C. Tetrahedron
Lett. 2001, 42, 7559-7561.
(17) Reed, E. A.; Curtiss, L. A.; Weinhold, F. Chem. Rev. 1988, 88,
899-926.
(18) (a) Valters, R. E.; Flitsch, W. Ring-Chain Tautomerism; Plenum
Press: New York, 1985. (b) Valters, R. E.; Fu¨lo¨p, F.; Korbonits, D.
Adv. Heterocycl. Chem. 1996, 66, 1.
(19) La´za´r, L.; Go¨blyo¨s, A.; Martinek, T. A.; Fu¨lo¨p, F. J . Org. Chem.
2002, 67, 4734-4741.
(20) Neuvonen, K.; Fu¨lo¨p, F.; Neuvonen, H.; Koch, A.; Kleinpeter,
E.; Pihlaja K. J . Org. Chem. 2001, 66, 4132-4140.
(21) Schuster, H.; Kathawala, F. G. Isoquinolines Bearing Basic Side
Chains. In The Chemistry of Heterocyclic Compounds, Vol. 38, Iso-
quinolines; Coppola, G. M., Schuster, H. F., Eds.; Wiley: New York,
1995; Part 3, p 1.
(22) (a) Hayashi, K.; Ozaki, Y.; Nunami, K.; Yoneda N. Chem.
Pharm. Bull. 1983, 31, 312-314. (b) Grunewald, G. L.; Sall, D. J .;
Monn, J . A. J . Med. Chem. 1988, 31, 824-830.
(26) Katritzky, A. R.; Suzuki, K.; He, H.-Y. J . Org. Chem. 2002, 67,
8224-8229.
5706 J . Org. Chem., Vol. 68, No. 14, 2003

Substituent-Dependent Negative Hyperconjugation
TABLE 1. P r op or tion s (%) of Ta u tom er ic F or m s in
Ta u tom er ic Equ ilibr ia for Com p ou n d s 9-14 (CDCl3, 300
K)
compd
X
σ+
9
11
12
10
13
14
a
b
c
d
e
f
g
h
i
p-NO2
p-CF3
p-Br
p-Cl
H
p-F
p-Me
p-OMe
p-NMe2 -1.7
0.79
0.612
0.15
0.114
0
3.4 88.7 7.9
4.8 88.2 7.0
7.6 86.1 6.3 14.7 28.9 56.4
6.9 88.0 5.1 8.1 38.1 53.8
9.0 86.3 4.7 12.8 28.1 59.1
9.6 85.4 5.0 9.9 38.6 51.5
4.1 25.8 70.1
4.7 32.0 63.3
-0.073
-0.311 11.6 85.0 3.4 14.3 28.9 56.8
-0.778 14.3 82.5 3.2 24.3 29.5 46.2
34.6 63.7 1.7 57.0 14.6 28.4
F IGURE 2. Final predominant minimum energy molecular
structures for 11e-14e, obtained by using ab initio HF/3-21G*
calculations.
calculated and the measured data. The conformational
search protocol comprised a stochastic search using the
Merck Molecular Force Field (MMFF94)27 and a subse-
quent minimization of the resulting low-energy confor-
mations at the ab initio level, using the HF/3-21G* basis
set for all the compounds 11(a -i)-14(a -i). The resulting
structures proved to be rigid, since no minor conformation
was found within the 6 kcal/mol energy window. The final
conformations for 11e-14e are shown in Figure 2.
The conformations obtained are supported by the NOE
patterns determined via the NOESY spectra. It is clear
from the structures that the ideal, antiperiplanar ar-
rangement cannot be found for the nitrogen lone pairs
and the possible endocyclic and exocyclic substituents on
C3.
NBO An a lysis. The exact analysis of the delocaliza-
tion energy contributions to the substituent-dependent
AE is a complex problem which can be tackled through
a second-order perturbative analysis of the Fock matrix
in the NBO basis. Although this procedure is known to
overestimate stabilizing interactions, they closely parallel
the energies afforded by the more accurate Fock matrix
deletion method.7b,28 The second-order energy stabiliza-
tion due to the electron donation from orbital i to orbital
j is given by eq 1, where qi is the donor orbital occupancy,
ꢀi and ꢀj are diagonal elements, and Fij is the off-diagonal
NBO Fock matrix element. The NBO calculations were
performed at the HF/6-31G* level using geometries
optimized at the HF/3-21G* level.
F IGURE 1. Plots of ∆G0 vs Hammett-Brown σ+ values for
epimerizations I (+) and II (×).
After the attainment of equilibrium, the specta of
compounds 9a -i or 10a -i contained well-separated
singlets resonating from the azomethine group at 8.31-
8.46 ppm or 8.23-8.44 ppm and the two N-CHAr-NH
hydrogens in the region 4.11-4.63 or 4.22-4.92 ppm.
With the aim of an unambiguous signal assignment,
standard 2D homo- and heteronuclear spectra were
recorded. Selected 1H chemical shifts for some charac-
teristic nuclei are presented in Table S1 (Supporting
Information). The proportions KX of the chain (9, 10) and
diastereomeric ring forms (11, 13 and 12, 14) of the
tautomeric equilibria 9a -i and 10a -i were determined
by integration of the well-separated NdCHAr (chain) and
N-CHAr-NH (ring) proton singlets in the 1H NMR
spectra (Table 1).
2
LFER analysis of the reaction free energies (Scheme
3, calculated at 300 K) demonstrates an acceptable linear
correlation with the Hammett-Brown substituent con-
stants (σ+) for epimerization I, while the linear fit falls
off for epimerization II (Figure 1).
Fij
Eij ) q
(1)
iꢀj - ꢀi
In the present model compounds, each nitrogen lone
pair can overlap with three different vicinal antibonding
orbitals associated with C3: namely σ*C3-N, σ*C3-Ar, and
σ*C3-H. (From here, the exact numbering of the model
ring system is followed.) Since we were interested in the
3-aryl substituent-dependent stereoelectronic interac-
tions, six orbital overlaps around C3 were taken into
consideration in our analysis. The results of the NBO
calculations revealed an insignificant amount of stabili-
zation energy for nN4-σ*C3-N and nN2-σ*C3-H in 11, for
nN4-σ*C3-N, nN4-σ*C3-H, nN2-σ*C3-N, and nN2-σ*C3-H in
The conformationally restrained condensed ring sys-
tems 11 and 12 display significant substituent effects on
the epimerization equilibrium, but the angularly con-
densed imidazoisoquinolines 13 and 14 do not exhibit net
substituent effects on the relative stabilities. Our conclu-
sion at this point is that the selected models are suitable
for detailed analysis of the stereoelectronic interactions
probably responsible for the substituent-dependent AE.
Con for m a tion a l An a lysis. Since the stereoelectronic
interactions are highly dependent on the geometry of the
studied molecules, thorough conformational analysis was
performed. Our goal was to determine the predominant
geometry for all the models and to identify the possible
minor conformers influencing the correlation between the
12, for nN4-σ*C3-N and nN2-σ*C3-H in 13, and for nN4
-
σ*C3-H and nN2-σ*C3-H in 14, which was accompanied by
(27) Halgren, T. A. J . Comput. Chem. 1996, 17, 490-519.
(28) Arnaud, R. J . Comput. Chem. 1994, 15, 1341-1356.
J . Org. Chem, Vol. 68, No. 14, 2003 5707

Hete´nyi et al.
F IGURE 5. Plots of theoretical stereoelectronic stabilization
energy ∆∑Eij vs Hammett-Brown σ+ values for epimerizations
I (+) and II (×).
F IGURE 3. Plots of theoretical stereoelectronic stabilization
energy ∆∑Eij vs ∆G0 values for epimerization I.
reaction free energy or on the calculated stereoelectronic
stabilization energy. These results raise two questions:
(i) Are the electron delocalization energies of the indi-
vidual orbital overlaps also substituent-dependent? (ii)
What specific NBO donor-acceptor interactions are
responsible for the substituent-dependent stabilization?
The Hammett-Brown substituent dependence of the
Eij values was tested by univariate linear regression for
both epimerization equilibria. The calculation was carried
out for each NBO donor-acceptor interaction, and the
summary of the analysis is displayed in Table 2; all the
linear fits exceeding a significance limit of 99% are given.
The results demonstrate that, regardless of the nature
of the antibonding orbitals associated with C3, the 3-aryl
substituent can have a significant influence on the two
orbital electron delocalization energies, even in com-
pounds 13 and 14. It is also seen that the σ+ values are
in an acceptably linear correlation with the Eij values,
despite the fact that the interacting orbitals are not in a
periplanar conformation. The evaluation of the slope of
the linear models (F) demonstrates that the 3-aryl sub-
stituents have a strong positive influence on the nN-
σ*C3-Ar interaction energies, i.e., the higher the value of
σ+, the higher the absolute value of the stabilization
energy. It is interesting that the interactions nN-σ*C3-N
and the nN-σ*C3-H change in an opposite sense: electron-
withdrawing substituents decrease the stabilizating de-
localization energy. The nN-σ*C3-H overlaps are subjected
to a weak modulation by σ+ as compared with those
F IGURE 4. Plots of theoretical stereoelectronic stabilization
energy ∆∑Eij vs ∆G0 values for epimerization II.
a negligible variation. The signed sum of the Eij values
gives the theoretical hyperconjugative stabilization en-
ergy ∆∑Eij in eq 2, where r and l designate the right-
hand side and the left-hand side, respectively, of the
epimerization equilibria.
∆
∑
E ) Er - El
ij
(2)
∑
∑
ij
ij
To correlate the experimental stability differences with
the theoretical results, linear regression analysis was
carried out with ∆∑Eij against ∆G0 (Figures 3 and 4).
For epimerization I, we obtained a good linear rela-
tionship, while in the case of epimerization II there was
no significant correlation. The variances of ∆G0 and ∆∑Eij
were of the same magnitude and the linear correlations
were acceptable for epimerization I. These findings
suggest that the observed 3-aryl substituent-dependent
stability difference is stereoelectronic in origin. The lack
of any linear relationship for epimerization II indicates
that the variations in ∆G0 cannot be explained by an
electron delocalization energy excess. The correlation
between ∆∑Eij and σ+ exhibits a similar pattern (Figure
5). The regression analysis for epimerization I gave a
linear fit, while ∆∑Eij for epimerization II proved inde-
pendent of σ+, which is in good agreement with the
experimental observations.
obtained for nN-σ*C3-Ar
.
Our observations reveal that all the possible vicinal
n-σ* interactions around C3 that have a conformational
arrangement other than orthogonal may contribute to the
substituent-dependent AE (Figure 6).
Translating the results of the NBO analysis to the
language of the valence bond theory gives an intuitive
picture of three types of double bond-no bond resonance
structures responsible for the phenomenological behavior
of the studied models (Scheme 4).
The growing electron demand of the 3-aryl substituent
increases the stabilizing contribution of the resonance
structure B, while the weights of the negative hypercon-
jugations C and D taper off. Depending on the exact
geometry of the species in the epimerization equilibrium,
the influence of the substituent on the donor-acceptor
interactions can result overall in an experimentally
observable substituent dependence of the equilibrium
It is clear that the experimentally detected substituent-
dependent stability difference for epimerization I is
connected with the AE comprising all the possible n-σ*
interactions around C3. For epimerization II, the elec-
tronic properties of the 3-aryl substituent do not seem
to exert a detectable effect either on the experimental
5708 J . Org. Chem., Vol. 68, No. 14, 2003

Substituent-Dependent Negative Hyperconjugation
TABLE 2. Su m m a r y of th e Lin ea r Regr ession An a lysis for th e Ha m m ett-Br ow n Su bstitu en t Dep en d en ce of th e
Hyp er con ju ga tive Sta biliza tion En er gy (Eij) Va lu es a n d Dih ed r a l An gles betw een th e In ter a ctin g Or bita ls (θ)
Eij
11
12
13
14
nN2-σ*C3-Ar
r
0.91
0.92
0.92
0.97
Fb (kcal/mol)
Cb (kcal/mol)
θ (deg)
-0.54 ( 0.09
-0.28 ( 0.04
-0.51 ( 0.08
-0.49 ( 0.04
-5.52 ( 0.07
-5.51 ( 0.03
-4.88 ( 0.06
-4.78 ( 0.03
131.49
0.94
-25.64
128.16
0.94
-26.20
nN2-σ*C3-N4
nN2-σ*C3-H
nN4-σ*C3-Ar
nN4-σ*C3-N2
nN4-σ*C3-H
r
a
a
a
0.97
Fb (kcal/mol)
Cb (kcal/mol)
θ (deg)
0.25 ( 0.04
0.28 ( 0.04
-1.13 ( 0.03
0.32 ( 0.03
-7.88 ( 0.02
-0.82 ( 0.03
-72.23
-83.15
a
a
a
147.29
0.85
68.12
a
a
a
6.82
0.93
146.02
a
a
a
84.74
0.97
r
a
a
a
F (kcal/mol)b
C (kcal/mol)b
θ (deg)
-10.42
0.93
r
F (kcal/mol)b
C (kcal/mol)b
θ (deg)
-0.42 ( 0.06
-0.45 ( 0.10
-7.97 ( 0.08
-0.43 ( 0.06
-4.62 ( 0.04
33.94
a
a
a
93.46
0.85
0.07 ( 0.02
-9.58 ( 0.01
156.99
-0.20 ( 0.02
-5.13 ( 0.05
-2.75 ( 0.01
-27.19
140.29
a
a
a
-80.16
a
a
a
38.53
0.92
r
a
a
a
F (kcal/mol)b
C (kcal/mol)b
θ (deg)
-0.05 ( 0.01
-9.45 ( 0.01
158.90
a
a
a
-86.11
0.89
0.13 ( 0.02
-8.32 ( 0.02
149.6
r
F (kcal/mol)b
C (kcal/mol)b
θ (deg)
-19.05
61.29
a
b
Significance level is below 99%. Regression was calculated for the linear model: Eij ) Fσ+ + C.
F IGURE 6. Graphical representations of the selected NBO interactions contributing to the substituent-dependent stereoelectronic
stabilization. The orbital overlaps are displayed for 11i.
SCHEME 4
orbital energy difference (∆ꢀ ) ꢀj - ꢀi) and the square of
the NBO Fock matrix off-diagonal element (Fij). ∆ꢀ-1
changes in parallel with σ+ for all types of acceptor
orbitals, and the relative increase in ∆ꢀ-1 on going from
p-NMe2 to p-NO2 is only a few percent (Figure 7).
This is in accordance with the picture of decreasing
acceptor orbital energy due to the increasing electron-
withdrawing effect of the 3-aryl substituent. Inspection
(29) The substituent effects of the 8,9-methoxy substituents in
compounds 10, 13, and 14 a priori cannot be ruled out. Since the
aromatic group is two bonds away from the C3-N4 bond, it might
perturb the polarization of this bond, and hence it may change the
stereoelectronic stabilization energy. Since the model compounds in
question are symmetric in that the aromatic substituent on C3 is also
at a distance of two bonds from the C10b-N4 bond, the substituent
effect in such cases can be estimated by using the results of our
calculations. For investigation of the influence on the polarization
coefficient of the C10b sp3 atomic orbitals Figure S1 was prepared. It
is seen that the polarization of σ*C10b-Ar and σ*C10b-N4 can be considered
to be independent from the σ+ values. This suggests that the methoxy
groups are unlikely to influence the substituent-dependent stereoelec-
tronic stabilization revealed by our analysis.
ratios (epimerization I), which may be interpreted as a
fine-tuned AE. It is also possible that the individual
hyperconjugative contributions cancel out, leading to an
equilibrium ratio that is independent of σ+ (epimerization
II).29
Equation 1 contains two important parameters deter-
mining the Eij values: the reciprocal donor-acceptor
J . Org. Chem, Vol. 68, No. 14, 2003 5709

Hete´nyi et al.
F IGURE 7. Plots of selected reciprocal donor-acceptor orbital
energy difference (∆ꢀ-1) in 11 obtained from NBO calculation.
The symbols 9, 2, and b designate the orbital overlaps nN2
σ*C3-N4, nN2-σ*C3-Ar, and nN2-σ*C3-H, respectively.
F IGURE 8. Plots of selected squared Fock matrix off-diagonal
elements (F2ij) in 11 obtained from NBO calculations. The
-
symbols 9, 2, and b designate the orbital overlaps nN2
σ*C3-N4, nN2-σ*C3-Ar, and nN2-σ*C3-H, respectively.
-
of the substituent dependence of F2ij (Figure 8) allows
the conclusion that the changes in Eij are mainly ex-
plained by the variations in Fij, since the changes are
higher as compared with those in ∆ꢀ-1, and the opposite
trend of the correlation with σ+ can also be detected. We
note that the Fock matrix off-diagonal element for the
nN-σ*C3-H overlap does not exhibit a good linear fit
against σ+.
Fij is strongly correlated with the orbital overlap matrix
element Sij, which finally boils down to a dihedral angle
and acceptor orbital polarization dependence.7b The sub-
stituent-dependent conformational changes for these
model compounds are around 1°, i.e., negligible. However,
the polarization coefficients of the C3 sp3 atomic orbitals
increase in σ*C3-Ar and decrease in the acceptor bonds
σ*C3-N and σ*C3-H when the electron demand of the 3-aryl
substituent is augmented (Figure 9).
Con clu sion
This extended view of the stereoelectronic interactions
affords a detailed description of the substituent-depend-
ent stability changes exhibited by the 2-aryl-1,3-X,N-
heterocycles. The electron-withdrawing properties of the
2-aryl substituent alter the polarization along all the
single bonds associated with C2; it therefore changes the
extent of the orbital overlaps between the nitrogen lone
pairs and the antibonding orbitals. The donor-acceptor
hyperconjugative stabilization energies display an ac-
ceptable linear correlation with the Hammett-Brown
substituent constant, and the theoretical stereoelectronic
stabilization energies calculated as the sum of the
individual orbital overlap electron delocalization energies
explain the experimentally observed substituent-depend-
ent AE.
To summarize, the 3-aryl substituent effect contracts
or expands the antibonding orbital in the direction of the
donor lone pairs (Figure 10), which accounts exactly for
the opposite trends of the variations in Fij, and thus for
the observed substituent dependence of the individual
orbital overlap stabilization energies.
Application of the results presented in this work may
help resolve the question of the substituent dependence
of the epimerization equilibrium constants in the field
of ring-chain tautomerism. In accord with the currently
increasing interest in dynamic covalent bonds allowing
the construction of supramolecular structures and com-
5710 J . Org. Chem., Vol. 68, No. 14, 2003

Substituent-Dependent Negative Hyperconjugation
recently recognized as useful building blocks for the
synthesis of novel supramolecular host compounds and
dynamic combinatorial libraries.31 It therefore appears
important to provide explanations as to the exact elec-
tronic mechanisms of the effects exerted by the substit-
uents and the stereochemistry on the equilibrium pro-
cesses of these compounds.
Exp er im en ta l Section
Compound 122a was prepared according to known proce-
dures.
1,2,3,4-Tetr a h yd r oisoqu in olin e-3-ca r boxa m id e (2). To
a solution of amino ester 1 (7.65 g, 40 mmol) in MeOH (25
mL) was added 25% methanolic ammonia solution (100 mL).
The mixture was allowed to stand in a well-closed container
at ambient temperature for 14 days, and the solvent was then
evaporated off. The crystalline product was filtered off, washed
with Et2O, dried, and recrystallized from EtOAc-EtOH: yield
1
5.78 g (82%); mp 164-166 °C (lit.22b mp 162-163 °C). The H
NMR data on the product correspond to the literature22b data.
3-Am in om eth yl-1,2,3,4-tetr a h yd r oisoqu in olin e Dih y-
d r och lor id e (3‚2HCl). To a stirred and cooled suspension of
LiAlH4 (2.85 g, 75 mmol) in dry THF (100 mL) was added
carboxamide 2 (4.41 g, 25 mmol) in small portions. The mixture
was stirred and refluxed for 3 h and then cooled, and the excess
of LiAlH4 was decomposed by the addition of a mixture of
water (5.7 mL) and THF (40 mL). The inorganic salts were
filtered off and washed with EtOAc (3 × 100 mL). The
combined organic filtrate and washings were dried (Na2SO4)
and evaporated under reduced pressure to give crude diamine
3 as an oil, which was converted to the crystalline dihydro-
chloride salt by treatment of its solution in MeOH with an
excess of 22% ethanolic HCl and Et2O. The crystalline dihy-
drochloride was filtered off, dried, and recrystallized from
MeOH-H2O-Et2O: yield 4.61 g (78%); mp 244-247 °C; 1H
NMR (D2O) δ 3.14 (dd, 1H, J ) 10.8, 17.1 Hz), 3.37 (dd, 1H, J
) 4.8, 16.4 Hz), 3.47 (dd, 1H, J ) 6.8, 13.9 Hz), 3.59 (dd, 1H,
J ) 6.3, 13.9 Hz), 4.1 (ddd, 1H, J ) 6.0, 11.3 Hz), 4.54 (s, 2H),
7.25-7.42 (m, 4H). Anal. Calcd for C10H16Cl2N2: C, 51.08; H,
6.86; N, 11.91. Found: C, 50.88; H, 6.61; N, 11.73.
r-Ben zyloxyca r b on yla m in o-N-[2-(3,4-d im et h oxyp h e-
n yl)eth yl]a ceta m id e (5). To a stirred and cooled (ice-salt
bath) solution of N-(benzyloxycarbonyl)glycine (4, 10.46 g, 50
mmol) and Et3N (5.06 g, 50 mol) in anhydrous toluene (100
mL) was added ethyl chloroformate (5.43 g, 50 mmol) dropwise
at a rate low enough to keep the internal temperature below
-10 °C. After 5 min, a solution of 2-(3,4-dimethoxyphenyl)-
ethylamine (9.06 g, 50 mmol) in dry CH2Cl2 (50 mL) was added
dropwise to the mixture, the internal temperature being kept
below 0 °C. When the addition was complete, the reaction
mixture was slowly heated to reflux and refluxed for 5 min.
The mixture was allowed to cool to room temperature and
washed with saturated aqueous NaHCO3 solution (2 × 50 mL)
and water (2 × 50 mL) after the addition of CHCl3 (200 mL).
The combined organic phases were dried (Na2SO4) and evapo-
rated in vacuo to give a crystalline product, which was filtered
off, washed with Et2O, dried, and recrystallized from EtOAc:
F IGURE 9. Substituent dependence of the polarization coef-
ficients in the C3-associated antibonding orbitals in 11. The
symbols 9, 2, and b designate the orbital overlaps σ*C3-N4
,
σ*C3-Ar, and σ*C3-H, respectively.
1
yield 11.55 g (62%); mp 122-124 °C; H NMR (CDCl3) δ 2.75
(t, 2H, J ) 7.1 Hz), 3.51 (dd, 2H, J ) 7.1, 13.1 Hz), 3.81 (d,
2H, J ) 5.5 Hz), 3.85 (s, 3H), 3.87 (s, 3H), 5.11 (s, 2H), 5.30
(br s, 1H), 5.93 (br s, 1H), 6.67-6.73 (m, 2H), 6.79 (d, 1H, J )
8.6 Hz), 7.30-7.39 (m, 5H). Anal. Calcd for C20H24N2O5: C,
64.50; H, 6.50; N, 7.52. Found: C, 64.23; H, 6.37; N, 7.48.
1-Ben zyloxyca r bon yla m in om eth yl-6,7-d im eth oxy-3,4-
d ih yd r oisoqu in olin e (6). To a stirred solution of compound
F IGURE 10. Graphical interpretation of the substituent
dependence of the polarization coefficients on C3. On going
from p-NMe2 to p-NO2, the C3-Ar antibonding orbital ex-
pands, while the C3-N and C3-H antibonding orbitals
contract at C3.
(30) Rowan, S. J .; Cantrill, S. J .; Cousins, G. R. L.; Sanders, J . K.
M.; Stoddart, J . F. Angew. Chem., Int. Ed. 2002, 41, 898-952.
(31) (a) Star, A.; Goldberg, I.; Fuchs, B. Angew. Chem., Int. Ed. 2000,
39, 2685-2689. (b) Star, A.; Fuchs, B. J . Org. Chem. 1999, 64, 1166-
1172.
binatorial libraries under equilibrium conditions,30 the
condensed 2-aryl-1,3-X,N-heterocycle derivatives were
J . Org. Chem, Vol. 68, No. 14, 2003 5711

Hete´nyi et al.
5 (11.17 g, 30 mmol) in dry CHCl3 (150 mL) was added POCl3
(13.80 g, 90 mmol). The mixture was refluxed for 3 h and then
evaporated. The oily residue was dissolved in hot water (200
mL), and the cooled solution was washed with EtOAc (2 × 50
mL). The aqueous phase was made alkaline with 20% NaOH
under cooling and extracted with CHCl3 (3 × 100 mL). The
combined organic extracts were dried (Na2SO4) and evaporated
in vacuo to give a crystalline product, which was filtered off,
washed with Et2O, dried, and recrystallized from MeOH: yield
7.55 g (71%); mp 123-126 °C; 1H NMR (CDCl3) δ 2.66 (t, 2H,
J ) 7.1 Hz), 3.69 (t, 2H, J ) 7.6 Hz), 3.89 (s, 3H), 3.92 (s, 3H),
4.39-4.44 (m, 2H), 5.15 (s, 2H), 6.31 (br s, 1H), 6.70 (s, 1H),
6.94 (s, 1H), 7.28-7.42 (m, 5H). Anal. Calcd for C20H22N2O4:
C, 67.78; H, 6.26; N, 7.90. Found: C, 67.99; H, 6.35; N, 7.68.
1-Be n zyloxyca r b on yla m in om e t h yl-6,7-d im e t h oxy-
1,2,3,4-tetr a h yd r oisoqu in olin e (7). To a stirred and ice-
cooled suspension of dihydroisoquinoline 6 (7.08 g, 20 mmol)
in MeOH (100 mL) was added NaBH4 (2.27 g, 60 mmol) in
small portions. The mixture was stirred for 3 h with cooling
and for 3 h without and then evaporated, and the residue was
dissolved in 5% HCl (150 mL). The solution was made alkaline
with 20% NaOH under ice cooling and extracted with CHCl3
(4 × 100 mL). The combined organic extracts were dried (Na2-
SO4) and evaporated in vacuo to give crude 7 as an oil, which
crystallized on treatment with n-hexane-Et2O. The crystals
were filtered off, washed with Et2O, and used in the next step
without further purification, yield 5.92 g (83%). For analytical
purposes, a small sample of 7 was recrystallized from Et2O:
Gen er a l P r oced u r e for th e P r ep a r a tion of 3-Ar yl-8,9-
dim eth oxy-1,2,3,5,6,10b-h exah ydr oim idazo[5,1-a ]isoqu in -
olin es (9, 11, 12a -i) a n d 3-Ar yl-1,2,3,5,10,10a -h exa h y-
dr oim idazo[1,5-b]isoqu in olin es (10, 13, 14a-i). To a solution
of diamine base 3 or 8 (3 mmol) in absolute MeOH (25 mL)
was added an equivalent amount of aromatic aldehyde (for
liquid aldehydes, a freshly distilled sample was used), and the
mixture was allowed to stand at ambient temperature for 1
h. The solvent was evaporated off and the oily product
crystallized on treatment with Et2O or n-hexane. The crystal-
line products 9a -i, 10a , and 10c-i were filtered off and
recrystallized. In the case of 10b, the evaporation was repeated
after the addition of toluene (10 mL), and the oily product was
dried in a vacuum desiccator for 24 h. The NMR spectrum
proved that the purity of this compound was greater than 95%.
All of the recrystallized new compounds (9a -i, 10a , and 10c-
i) gave satisfactory data on elemental analysis (C, H, N (
0.3%). The physical data on compounds 9 and 10 are listed in
Table S8 in the Supporting Information.
Com p u ta tion a l Meth od ology. The conformational search
protocol comprised a stochastic search using MMFF94 imple-
mented in the Chemical Computing Group’s MOE software.
The ab initio calculations were carried out with Gaussian 94
at the HF/3-21G*//HF/6-31G* level. The NBO analysis was
performed by using the NBO 3.1 software implemented in
Gaussian 94 with default parameters. The overlapping NBO
orbitals were visualized with the molecular graphics package
MOLEKEL.32
1
mp 93-95 °C; H NMR (CDCl3) δ 2.64-2.70 (m, 2H), 2.94-
3.02 (m, 1H), 3.03-3.13 (m, 1H), 3.25-3.34 (m, 1H), 3.62-
3.70 (m, 1H), 3.85 (s, 6H), 3.96-4.03 (m, 1H), 5.11 (s, 2H), 5.45
(br s, 1H), 6.57 (s, 1H), 6.64 (s, 1H), 7.28-7.38 (m, 5H). Anal.
Calcd for C20H24N2O4: C, 67.40; H, 6.79; N, 7.86. Found: C,
67.13; H, 6.58; N, 7.61.
Ack n ow led gm en t. We thank the Hungarian Re-
search Foundation (OTKA No. TS 040888) for financial
support.
1-Am in om et h yl-6,7-d im et h oxy-1,2,3,4-t et r a h yd r oiso-
qu in olin e d ih yd r obr om id e (8‚2HBr ). Compound 7 (3.56 g,
10 mmol) was suspended in 33% hydrobromic acid in acetic
acid (15 mL), and the mixture was heated gently with
occasional shaking until all of the substance had been dis-
solved. The mixture was allowed to stand at ambient temper-
ature for 30 min, and Et2O (30 mL) was then added. The
crystalline dihydrobromide salt of 8 that formed was filtered
off, washed with a mixture of MeOH and Et2O, dried, and
recrystallized from MeOH-H2O-Et2O: yield 2.99 g (78%); mp
268-269 °C; 1H NMR (D2O) δ 3.10 (t, 2H, J ) 6.6 Hz), 3.52-
3.70 (m, 3H), 3.75 (dd, 1H, J ) 8.1, 14.6 Hz), 3.87 (s, 6H), 4.91
(dd, 1H, J ) 4.5, 8.1 Hz), 6.92 (s, 1H), 6.97 (s, 1H). Anal. Calcd
for C12H20Br2N2O2: C, 37.52; H, 5.25; N, 7.29. Found: C, 37.27;
H, 5.06; N, 7.39.
Su p p or tin g In for m a tion Ava ila ble: Cartesian coordi-
nates and HF energies for 11-14. Donor-acceptor stabiliza-
tion energies obtained from the second-order perturbation
theory analysis of the Fock matrix in the NBO basis. The off-
diagonal Fock matrix elements in the NBO basis. Donor-
acceptor orbital energy differences. Polarization coefficients of
the C3-associated antibonding orbitals. Physical data on
compounds 9, 11, 12 and 10, 13, and 14. Selected characteristic
1H chemical shifts (ppm, δTMS ) 0 ppm) for Compounds 9a -
14a . Experimental free energies and theoretical stabilization
energies for epimerizations I and II. Substituent dependence
of the polarization coefficients in the C10b-associated anti-
bonding orbitals in 13. This material is available free of charge
via the Internet at http://pubs.acs.org.
Pure diamine bases 3 and 8 were obtained from the above
dihydrohalide salts by alkaline treatment (20% NaOH), ex-
traction (CHCl3), and evaporation under reduced pressure. The
free bases were dried in a vacuum desiccator for 24 h before
further transformations.
J O034417+
(32) Portmann, S.; Lu¨thi, H. P. Chimia 2000, 54, 766-770.
5712 J . Org. Chem., Vol. 68, No. 14, 2003