Conformational complexity in seven-membered cyclic triazepinone/open hydrazones. 1. 1D and 2D variable temperature NMR study

Article abstract of DOI:10.1021/jo962322k

The stereochemistry and conformational behavior of a series of 22 2- methyl-2-alkyl(phenyl, aryl)-4-N-methyl-1,2,3,4-tetrahydro-5H-1,3,4- benzotriazepin-5-ones and their open-chain hydrazone tautomers in various solvents were studied by 1D and 2D NMR tech

Full text of DOI:10.1021/jo962322k

5080
J . Org. Chem. 1997, 62, 5080-5088
Con for m a tion a l Com p lexity in Seven -Mem ber ed Cyclic
Tr ia zep in on e/Op en Hyd r a zon es. 1. 1D a n d 2D Va r ia ble
Tem p er a tu r e NMR Stu d y
Kalevi Pihlaja,* Mario F. Simeonov,† and Ferenc Fu¨lo¨p‡
Department of Chemistry, University of Turku, FIN-20014 Turku, Finland
Received December 11, 1996X
The stereochemistry and conformational behavior of a series of 22 2-methyl-2-alkyl(phenyl, aryl)-
4-N-methyl-1,2,3,4-tetrahydro-5H-1,3,4-benzotriazepin-5-ones and their open-chain hydrazone
tautomers in various solvents were studied by 1D and 2D NMR techniques in the temperature
range from 193 K to 410 K. Molecular rearrangements involving interconversions of the ring and
open-chain forms (the latter via amide bond and Z/ E CdN double bond isomerization), pseudoro-
tation of the ring forms, and N-inversion processes with different rates on the NMR time scale
1
took place, leading to the observation of average and deceptively simple H and 13C NMR spectra
for most of them at room temperature.
Sch em e 1. Com p ou n d s Stu d ied
In tr od u ction
The therapeutic and commercial success of 1,4-benzo-
diazepines has stimulated considerable interest in related
systems.1 Since the biological activity is dependent on
the stereochemistry of the seven-membered ring,2 we
were interested in studying a series of variably substi-
tuted 1,3,4-benzotriazepinones. Considerable confusion
has surrounded the preparation3 of these 1,4-benzodiaz-
epine analogues due to the tautomeric process.4 Many
of the early benzotriazepinone structures were reassigned
later as quinazolinones, imidazolines, oxadiazoles, and
other structures,3 i.e. the reactions of â-aminocarbohy-
drazides with oxo reagents can result in hydrazone,
pyrimidinone, or triazepinone derivatives, depending on
the substituents and the conditions employed.5
The first example of the formation of a seven-mem-
bered ring from an o-amino-substituted aromatic hy-
drazide was the reaction of anthranilic acid N,N′-
dimethylhydrazide with carbonyl compounds.5a Later,
the rearrangements of o-aminobenzoyl N-methylhydra-
zones of â-dicarbonyl5b and simple monocarbonyl5c com-
pounds were shown by NMR spectroscopy to involve a
ring/open-chain equilibrium in solution. N-Methyl car-
bohydrazides were observed to react with acetaldehyde,
acetone, and cyclohexanone to yield cyclic products
(condensed triazepinones), with benzaldehyde5c to yield
open forms (hydrazones), and with acetophenone to yield
† Institute of Organic Chemistry with Center of Phytochemistry,
Bulgarian Academy of Sciences, BG-1113 Sofia, Bulgaria.
‡ Institute of Pharmaceutical Chemistry, Albert Szent-Gyo¨rgyi
Medical University, H-6720 Szeged, POB 121, Hungary.
X Abstract published in Advance ACS Abstracts, J uly 1, 1997.
(1) Bachman, G. B.; Levine, H. A. J . Am. Chem. Soc. 1947, 69, 2341.
(2) Fryer, R. I.; Cook, C.; Gilman, N. W.; Walser, A. Life Sci. 1986,
39, 1947
(3) (a) Sunder, S.; Peet, N. P.; Trepanier, D. L. J . Org. Chem. 1976,
41, 2732. (b) Peet, N. P.; Sunders, S. J . Heterocycl. Chem. 1984, 21,
1807. (c) Scheiner, P.; Frank, L.; Giusti, I.; Arwin, S.; Pearson, S. A.;
Excellent, F.; Harper, A. P. J . Heterocycl. Chem. 1984, 21, 1817. (d)
Leiby, R. W. J . Heterocycl. Chem. 1984, 21, 1825. (e) Peet, N. P.
Synthesis 1984, 1065. (f) Fu¨lo¨p, F.; Simeonov, M.; Pihlaja, K. Tetra-
hedron 1992, 48, 531.
(4) Valters, R.; Fulo¨p, F.; Korbonits, D. Adv. Heterocycl. Chem. 1996,
68, 1.
(5) (a) Hromatka, O.; Knollmu¨ller, M.; Krenmu¨ller, F. Monatsh.
Chem. 1969, 100, 941. (b) Ga´l, M.; Fehe´r, O.; Tihanyi, E.; Horva´th,
Gy.; J erkovich, Gy. Tetrahedron 1982, 38, 2933. (c) Ga´l, M.; Tihanyi,
E.; Dvortsak, P. Acta Chim. Hung. 1986, 123, 55. (d) Reddy, P. P.;
Reddy, C. K.; Reddy, P. S. N. Bull. Chem. Soc. J pn. 1986, 59, 1575.
a hydrazone benzotriazepinone ring/open-chain equilib-
rium, with tautomerism, in a 1:1 ratio (Scheme 1).
S0022-3263(96)02322-5 CCC: $14.00 © 1997 American Chemical Society

Conformations of Cyclic Triazepinone/Open Hydrazones
J . Org. Chem., Vol. 62, No. 15, 1997 5081
Ta ble 1. 1H a n d 13C NMR Ch em ica l Sh iftsa (p p m ), P op u la tion s (%), a n d Th er m od yn a m ic P a r a m eter s (k ca l m ol-1) of
Com p ou n d 1
δ, 293 K
δ, 273 K
δ, 233 K
nucleus
ring form
open form
ring form
open form
ring form
(C-2)-CH3
(C-2)-CH3
(N-4)-CH3
(N-1)-H
1.37b (27.2)
1.37b (27.2)
3.30 (38.8)
4.48
1.84b (19.6)
2.06b (24.9)
3.23 (nd)c
4.48
1.29b (nd)
1.40b (nd)
3.27 (nd)
4.54
1.84b (nd)
2.06b (nd)
3.27 (nd)
4.54
1.20 (27.0)
1.45 (26.7)
3.29 (38.6)
4.60
(N-3)-H
3.73b
-
3.90b
-
4.00b
population
98
2
99
1
100
∆∆G°(ringhopen)
∆∆G°(ring dynamic process)
2.3
>2.5
13.0
other nuclei
C-2
C-5
(77.4)
(172.7)
(nd)
(161.7)
(nd)
(nd)
(77.6)
(172.7)
a
b
The latter in parentheses. Signal broadening. c nd ) not detected.
Factors such as N-substitution, ring-closure, possibility
of conjugation, and steric hindrance have been found to
govern the equilibria.5c Thus, it is evident that, like other
seven-membered benzoheterocyclic compounds,4 the sub-
stituted benzotriazepines are extremely complex and
rather flexible systems. Not surprisingly, therefore, little
is known about the conformations attained by these
systems in solution.6a Actually, only a few examples of
1,3-N,N-heterocycles showing ring/open-chain tautomer-
ism are known.4,6b-d
The primary goal of this work was to determine the
structure in solution of the studied compounds. However,
a detailed conformational analysis was required due to
the wealth of dynamic chemistry that embodies not only
the ring/open-chain rearrangement, but also the inter-
conversions of the ring and open-chain forms (the latter
via an amide bond and Z/ E CdN double bond isomer-
ism), pseudorotation of the ring, and N-inversion pro-
ceeding with different rates on the NMR time scale, but
leading to the average and deceptively simple 1H and 13C
NMR spectra at room temperature.
All these findings impelled us to set two goals for this
initial work: (i) by means of 1D and 2D NMR studies at
variable temperatures, to characterize the complex con-
formational properties of compounds 1-22 (Scheme 1)
in different solvents (this work); and (ii) to relate the
NMR thermodynamic parameters to molecular model
calculations and X-ray measurements in order to gain
additional insight into the dynamic processes in these
seven-membered benzoheterocycles (Part 26e). A detailed
computational study of the open-chain hydrazone forms
found for most of the compounds will be reported else-
where.
F igu r e 1. 1H NMR spectra of 1 in CD2Cl2 at (a) 293, (b) 273,
and (c) 233 K.
In CD2Cl2 at room temperature (293 K), 1 is present
1
in two forms: R and O1. Crucial parts of the H NMR
spectrum of 1 at different temperatures are shown in
Figure 1. The spectrum at 293 K exhibits broad signals
at 1.37 and 3.73 ppm, corresponding to the (C-2)-methyl
protons and (N-3)-H proton of the ring form, respectively.
The former peak correlates with the peak at 77.4 ppm
(C-2) in the 2D LR HETCORR spectrum at 293 K. The
13C spectrum (Figure 2a) also reveals some weak signals
due to the open form, e.g. C-2 in this form in 2-4
resonates at ca. 170 ppm (see Table 2; the NMR param-
eters for 5 and 6 are very close to those of compound 4).
Resu lts a n d Discu ssion
Com p ou n d s 1-6. Rin g F or m s. The 1H and 13C
NMR chemical shifts of 1 and 2-4 are listed in Tables 1
and 2, respectively. The cross-peaks observed in the 2D
HETCORR and 2D LR HETCORR NMR spectra were
crucial for the complete assignment of the spectra.
Further, the special changes occurring upon change of
the solvent and during thermal activation permitted
assignment of the different sets of peaks up to three [one
ring (R) and two open-chain (O1 and O2)] forms.
1
Further, the H NMR signals of 1 at 3.30 and 3.23 ppm
were assigned to the (N-4)-methyl protons of the R and
O1 isomers, respectively. The signal at 3.30 ppm has
cross-peaks with the carbon resonances at 38.8 (in the
2D HETCORR spectrum) and 172.7 ppm (in the 2D LR
HETCORR spectrum), assigned to the (N-4)-methyl and
(6) (a) Messinger, J .; Buss, V. J . Org. Chem. 1992, 57, 3320. (b)
Korbonits, D.; Tobia´s-He´ja, E.; Kolonits, P. Chem. Ber. 1991, 124, 1199.
(c) Korbonits, D.; Tobia´s-He´ja, E.; Simon, K.; Kolonits, P. Chem. Ber.
1991, 124, 2065. (d) Lambert, J . B.; Wang, G.; Huseland, D. E.; Takiff,
L. C. J . Org. Chem. 1987, 52, 68. (e) Simeonov, M. F.; Fu¨lo¨p, F.;
Sillanpa¨a¨, R.; Pihlaja, K. J . Org. Chem. 1997, 62, 5089.
1
C-5 of the ring form, respectively. The H and 13C NMR

5082 J . Org. Chem., Vol. 62, No. 15, 1997
Pihlaja et al.
F igu r e 2. 13C NMR spectra of 1 in CD2Cl2 at (a) 293 and (b)
233 K.
data for the ring conformer of 1 agree closely with those
observed by Ga´l et al.7 All conformational energy differ-
ences (∆G°) between the ring and open-chain forms were
estimated from the integrals (I) of the (N-4)-methyl and/
or (C-2)-methyl protons, using the equation ∆G° ) -RT
ln K, where K ) I(major form)/I(minor form)8 for 1-4 (see
Tables 1 and 2).
When a CD2Cl2 solution of 1 is cooled, the 13C NMR
spectrum at 233 K (Figure 2b) reveals only one set (R-
form) of carbon signals. Now two peaks relating to the
(C-2)-methyl carbon, at 27.0 ppm and 26.7 ppm, and a
strong C-2 resonance at 77.6 ppm are observed. In the
1H spectra of 1 at 273 K and 233 K (Figure 1b,c), the
peak at 1.37 ppm is split into a doublet (1.20 ppm and
1.45 ppm), whereas the signal of the (N-3)-H protons at
4.00 ppm is still a broad singlet. Thus, within the
temperature range examined, three dynamic processes
occur in 1: (i) a rapid ring interconversion affecting
mainly N-3 and C-2 and the respective methyl groups
with an energy barrier (∆Gq value, calculated via the
coalescence temperature method8) of ca. 13.0 kcal mol-1
at Tc ) 273 K and freezing out at 233 K; (ii) ring/open-
chain rearrangement with ∆Gq higher than 13.0 kcal
mol-1 and with a Gibbs energy difference (∆G°) of 2.3
kcal mol-1 in favor of the ring form; and (iii) a rapid (on
the NMR time scale) N-3 nitrogen inversion process with
a ∆Gq less than 13.0 kcal mol-1
.
1
Nearly the same H and 13C NMR chemical shifts as
those for 1 were observed for 2, except for the downfield
shifted C-2 signal (of the ring form) by ca. 10 ppm. The
latter must be due to the electron-withdrawing effect of
the chlorine atom on C-7, an evidence for the para-para
interaction in the (C-9a)-(N-1)-(C-2) moiety, as sug-
gested by the X-ray results and molecular modeling. The
ring/open and open/open molecular rearrangements of 2
appear to be much slower than those of 1, since the
proportion of the open form was still increasing after a
week, and peaks corresponding to other open forms were
also appearing [mainly for the (N-4)-methyl protons; see
Table 2].
(7) Ga´l, M.; Pallagi, I.; Soha´r, P.; Fu¨lo¨p, F.; Ka´lma´n, A. Tetrahedron
1989, 45, 3513.
(8) Rahman, Atta-ur. Nuclear Magnetic Resonance, Basic Principles;
Springer Verlag: New York, 1985.

Conformations of Cyclic Triazepinone/Open Hydrazones
J . Org. Chem., Vol. 62, No. 15, 1997 5083
1
and the bulky alkyl groups. Our 1D H NOE difference
measurements on 4 show that (i) irradiation of the tert-
butyl protons (at 1.03-1.05 ppm, ring/open forms) gives
NOE (14%) at 1.77 ppm [(C-2)-methyl protons, O1 form,
see below) and some effects at 7.19 and 6.54 ppm
(aromatic H-8 and H-9 of the ring form, respectively), but
not at the (N-4)-methyl protons, in agreement with the
inverted boat form, and (ii) irradiation of the ring (C-2)-
methyl protons at 1.34 ppm gives NOE at 3.30 ppm [(N-
4)-methyl protons of the ring form] and on the tert-butyl
protons in agreement with the above assumptions.
On increase of the temperature from 293 to 323 K the
population of the ring form of 3 decreased from ca. 70%
to 62%, mainly at the expense of the major open form.
Heating a sample of 4 in DMSO-d6 caused more drastic
changes in the population ratios, and only one set of
signals for open forms was detected at 433 K (cf. Table
2).
As for 2, the time-dependent 1H NMR spectrum of 4
provides an insight into the thermodynamic features of
the ring/open-chain equilibria: initially, the ring/open-
chain ratio is ca. 1:4; after a week, the peaks of the open
form became broad; later, peaks due to the second open-
chain form were detected. Thus, one can speculate that
in the crystalline state 4 attains the open form, that the
energy barrier of the ring/open-chain interconversion of
this compound is less than that between the open forms,
and that the latter in turn is less than the energy needed
for the ring transformation. Similarly, it can be esti-
mated that the energy barriers of the ring/open-chain and
ring-inversion processes for 4 are higher than those for
1-3. The observed difference (ca. 1.5 kcal mol) between
the values of ∆G° (ring/open) for 1 and 3, and for 3 and
4, can be explained simply by the change in the total
nonbonded through-space methyl/methyl interactions on
going from 1 to 3, and from 3 to 4.
For 5, a small change in the ring/open-chain ratio was
observed, which can be attributed to the electronic effect
of the (C-7)-chlorine, as in the case of 2. The NMR
spectral parameters of 6 are similar to those for 1 (i.e.
the conformational behavior is related), besides the fact
that no signal broadening was found, except that of the
(N-3)-H proton, due to a nitrogen-inversion process.
Op en -Ch a in F or m s (h yd r a zon es). As mentioned
above, one or two sets of signals relating to open forms
F igu r e 3. 1H and 13C NMR spectra as a function of solvent
and temperature for 4: (a) 1H NMR spectrum in CDCl3 at 293
K, (b) 13C NMR spectrum in CDCl3 at 293 K; (c) 1H NMR
spectrum in DMSO-d6 at 293 K; (d) 1H NMR spectrum in
DMSO-d6 at 433 K; (e) 13C NMR spectrum in DMSO-d6 at 433
K.
The 1H and 13C NMR spectra of 3 and 4 are rather
complex. They were interpreted in terms of a superposi-
tion of three subspectra corresponding to one ring and
two open forms, the latter being favored for 4 (see Table
2). The 1H and 13C NMR spectra for 4 are shown in
Figure 3. The complete assignment was achieved through
analysis of the 2D HETCORR and 2D LR HETCORR
NMR spectra. A similar broadening as for 1 was ob-
served, i.e. the same dynamic processes must occur in 3
and 4. An upfield shift of the (C-2)-methyl carbon signal
occurred, from ca. 27 ppm in 1 to ca. 20 ppm in 3 and 4,
which must be due to the well-known steric compression
shift effect9,10 of the bulky substituents in 3 and 4. A
similar shift was not observed for the (N-4)-methyl
carbon. Thus, it may be concluded that the favored ring
form of 3 and 4 in solution is an inverted boat, with the
(C-2)-methyl group orientated between the (N-4)-methyl
1
were observed in the H and 13C NMR spectra of 1-4.
The assigned chemical shifts of these isomers are given
in Tables 1 and 2. C-2, which resonates at ca. 80 ppm
in the ring form, is deshielded to ca. 165 ppm in the open-
chain hydrazones. The chemical shifts of the (C-2)-
azomethine carbon of hydrazones are known to be
between ca. 150 ppm and 180 ppm,10 while those for the
hydrazones of aliphatic aldehydes and alicyclic ketones
are ca 180 ppm and those for the corresponding aromatic
aldehydes ca 150 ppm.11 In some pyruvate hydrazones,
the greater electron-withdrawing power of pyruvate as
compared with that of aryl groups causes additional
upfield shifts to ca. 140 ppm.12 The same effect has been
observed in some arylaldehyde N,N-dialkylhydrazones.13
1
Two sets of H and 13C chemical shifts were observed
for the (C-2)-methyl group in the hydrazones of 1-6
(9) Grant, D. M.; Cheney, B. V. J . Am. Chem. Soc. 1967, 89, 5315.
(10) Breitmaier, E.; Voelter, W. NMR Spectroscopy; Verlag Che-
mie: Weinheim, 1974. Kalinowksi, H.; Berger, S.; Braun, S. NMR
Spektroscopie; Georg Thieme Verlag: Stuttgart, New York, 1984.
Pihlaja, K.; Kleinpeter, E. Carbon-13 NMR Chemical Shifts in Struc-
tural and Stereochemical Analysis; VCH Publishers, Inc.; New York,
1994.
(11) Fu¨lo¨p, F.; Semega, EÄ .; Berna´th, G.; Soha´r, P. J . Heterocycl.
Chem. 1990, 27, 957.
(12) Palla, G.; Predieri, G.; Domiano, P.: Vignali, C.; Turner, W.
Tetrahedron 1986, 42, 3649.
(13) Brehme, R.; Radeglia, R.; Gru¨ndemann, E. J . Prakt. Chem.
1990, 332, 911. Brehme, R. Chem. Ber. 1990, 123, 2039. Brehme, R.;
Klemann, A. Tetrahedron 1987, 43, 4113.

5084 J . Org. Chem., Vol. 62, No. 15, 1997
Pihlaja et al.
Sch em e 2
(Tables 1 and 2): (i) at ca. 1.8 and ca. 20 ppm, respec-
tively (at higher field in the cases of 3 and 4), for the
major form (O1); and (ii) at ca. 2.2 and ca. 24 ppm,
respectively, for the minor form (O2). Similar values
were observed by Palla et al.12 for some N-acyl- and
N-aroylhydrazones. In principle, two Z/ E geometrical
isomers and two cis/ trans-amide conformers can be found
for the studied hydrazones as concerns the azomethine
double bond and amide bond isomerism, respectively
(Scheme 2).
For 1, it is reasonable to assign the 1H and 13C chemical
shifts of 1.84 and 19.6 ppm, respectively, to the (C-2)-
methyl group closer to the (N-4)-methyl protons and thus
upfield shifted due to the steric compression shift, i.e.
γ-gauche effect.9,10 Thus, the major open-chain form (O1)
in 3-6 corresponds to the E isomer and the minor form
(O2) to the Z isomer.
Additionally, the literature data12 show that the ben-
zoylhydrazone of methyl pyruvate, both in DMSO-d6 and
in CDCl3, prefers the trans-amide conformer, as in the
solid state. Our NMR data also suggest that in the open
forms the rotation around the (C-5)-(C-5a) bond is
hindered by the intramolecular H-bond of type N-H‚‚‚O(C)
as evidenced by the upfield chemical shift of the phenyl
proton in the ortho position to the carbonyl group,
whereas in the ring forms of 1-20 these signals appear
at ca. 7.8 ppm, separately from the multiplet of the other
aromatic protons at 6.6-7.3 ppm.
F igu r e 4. 1H NMR spectrum of 7 in CD2Cl2 at (a) 293, (b)
273, (c) 253, (d) 233, (e) 213, and (f) 193 K.
in some N,N-dimethyl- and diphenylhydrazones,15 which
are high as compared to our observations on the time-
dependent changes in 2-6: the signal broadening for the
(C-2)- and (N-4)-methyl protons in CD2Cl2 at room
temperature, and the calculated energy barrier of ca. 18
kcal mol-1 for 4 in DMSO-d6 at 323 K. On the other
hand, Palla et al.12 showed that the electronic effects can
change the energy barriers of the cis/ trans-amide bond
and/or Z/ E processes, depending on the substituents (see
1
the discussion below). Because of the lack of definite H
and 13C NMR evidence for the structural assignment in
solution [e.g. from the chemical shifts of the (N-4)-methyl
and (C-2)-methyl groups in the different isomers] and
because of the conformationally relevant NOE values,
which can be misleading in case of molecular equilibra-
tions which proceed at rates fast on the NMR scale,16 our
considerations should also be treated with some caution.
In order to investigate the cis/ trans and Z/ E preferences,
and also the phenomenon of N-N rotation as a function
of structure, MM and semiempirical MO calculations are
in progress for these and other model compounds.
Com p ou n d 7. This compound belongs in the series
(7-20) with a phenyl or aryl group at C-2. At room
temperature, all these molecules seem to be even more
flexible than 1-6 and to undergo the same complex
rearrangement as discussed above for 1-6. Therefore,
The question of the preference of a trans- or cis-amide
structure as a consequence of N-methylation is quite
interesting and remains open for some compounds.14 Our
1
a detailed H and 13C NMR study of 7 was undertaken
1
1D H NOE difference measurements on 4, immediately
in the temperature range from 190 K to 390 K in CD2Cl2
and DMSO-d6 as solvents, and time-dependent spectra
were also recorded.
after dissolution, when the O1 form predominates (see
above), show a strong (ca. 36%) NOE effect at 7.16 ppm
and ca. 7% NOE at 6.67 ppm (aromatic H-6 and H-7,
respectively, in an open form) when the N-methyl protons
are irradiated. This result supports the trans-amide
bond arrangement in O1.
Immediately after dissolution, the 1H and 13C NMR
spectra of 7 at 293 K can be interpreted in terms of two
superimposed subspectra corresponding to (i) averaged
broad and weak signals of the ring forms and (ii) another
set of sharper peaks relating to the open-chain isomer-
(s) (Figures 4a and 5a, respectively) and in a ca. 1:1 ratio.
All the above experimental findings prompt us to
believe that the hydrazones of 1-6 are (i) in a trans-
amide configuration in the crystalline state and in
solution; (ii) the thermal, time- and solvent-induced open-
chain isomerization of 1-6 discussed above involves their
Z/ E transformation. The literature data show that the
energy barrier for this process is ca. 25 kcal mol-1 in some
N-acyl- and N-aroylhydrazones12 and ca. 23 kcal mol-1
1
Comparison of the H spectrum of 7 with those of 1-6
led to the postulation that the broad peak at 2.7 ppm
belongs to the (N-4)-methyl protons of the ring form,
which therefore seem to be located in the shielding cone
of the (C-2)-benzene ring. Thus, the most upfield, broad
(15) Shvo, Y.; Nahlieli, A. Tetrahedron Lett. 1970, 4273.
(16) Kessler, H.; Griesinger, C.; Lautz, J .; Mu¨ller, A.; van Gunsteren,
W. F.; Berendsen, H. J . C. J . Am. Chem. Soc. 1988, 110, 3393.
(14) Itai, A.; Toriumi, Y.; Saito, S.; Kagechika, H.; Shudo, K. J . Am.
Chem. Soc. 1992, 114, 10649.

Conformations of Cyclic Triazepinone/Open Hydrazones
J . Org. Chem., Vol. 62, No. 15, 1997 5085
1
The full H and 13C NMR assignments of 7 in CD2Cl2 at
293 K are given in Table 4.
At low temperature (see Figures 4b-f and 5b-f), both
the 1H and 13C NMR spectra exhibit more complex
patterns. Thus, on cooling to 273 K, the broad peaks at
2.7 and 4.2 ppm [(N-3)-H] were first sharpened and
shifted downfield (i.e. N-inversion becomes slower) and
the peak at 1.7 ppm was split into two singlets (1.62 and
1.79 ppm). Additionally, a peak appeared at 3.46 ppm,
due to the (N-4)-methyl protons not located in the
shielding cone of the (C-2)-aryl group. The coalescence
of this ring inversion process occurred near room tem-
perature, and the energy of activation was calculated to
be ca. 14.7 kcal mol-1. On further cooling to below 253
K, a clear splitting of the signals at 2.3 and 3.4 ppm
occurred, due to the freezing of the interconversion of the
open-chain forms to trans-E (O1) and cis-E (O2), as
confirmed by the 13C spectra (see below), and the energy
of activation was calculated to be ca. 12.8 kcal mol-1 (Tc
) 253 K). These conclusions were also consistent with
the relative intensities observed for the two subspectra
of the ring and open-chain isomers: the changes in peak
intensity at 1.7 ppm are reflected in the peak intensity
at 2.7 ppm, and those at 2.3 ppm in the peak intensity
at 3.4 ppm.
F igu r e 5. 13C NMR spectrum of 7 in CD2Cl2 at (a) 293, (b)
In the 13C NMR spectrum of 7 at 253 K, two sets of
peaks corresponding to the two ring conformations were
distinguished, the most characteristic of them being those
at 28.1 ppm [(C-2)-methyl carbon], 37.5 ppm and 39.0
ppm [(N-4)-methyl carbon], 81.1 ppm and 81.9 ppm (C-
2), 142.6 ppm and 143.4 ppm (C-9a) (see Table 4).
Further, at 233 K the ring forms gave two more signals,
at 172.1 ppm and 172.2 ppm (C-5), whereas the peaks of
the open forms, after some broadening at 253 K (near
coalescence), started to split at this and lower tempera-
tures: the C-2 signal at 170.1 and 169.7 ppm (233 K);
the C-9a signal at 146.7 and 147.0 ppm (233 K); the
N-methyl carbon at 35.9 and 41.0 ppm (233 K); the (C-
2)-methyl carbon at 17.1 ppm split into two peaks (16.5
and 17.2 ppm) at 213 K, etc. (see Table 4). Below 213 K,
the peaks at 167.0 and 168.5 ppm were assigned to C-5
in the open forms, i.e. they are shifted about 10 ppm
upfield in comparison with those for the open forms of
1-6.
273, (c) 253, (d) 233, (e) 213, and (f) 193 K.
Ta ble 3. 1H a n d 13C NMR Ch em ica l Sh iftsa (p p m ) of
Com p ou n d 7 in CDCl3 a t 293 K
nucleus
ring form
open form
(C-2)-CH3
(N-4)-CH3
(N-1)-H
(N-3)-H
C-2
1.70b (28.6b)
2.70b (37.9b)
4.70b
2.26 (17.1)
3.38 (37.9b)
4.80
4.20b
-
(81.2)
(170.5)
C-5
(172.4)
(168.0)
(118.2)
C-5a
(128.3)
HC-6
HC-7
HC-8
HC-9
7.84 (131.7)
7.03 (121.2)
7.33 (132.5)
6.80 (120.8)
(143.2)
7.22 (129.4)
6.65 (116.6)
7.15 (131.1)
6.70 (116.8)
(146.8)
C-9a
C-1′
(131.8)
(137.2)
HC-2′,6′
HC-3′,5′
HC-4′
7.30 (128.4)
7.70 (127.1)
7.40 (130.7)
7.30 (128.0)
7.75 (127.1)
7.40 (130.7)
a
b
The results from the 2D HETCORR and the optimized
2D LR HETCORR experiments at 193 K were also
consistent with our conclusions. Thus, in the 2D HET-
CORR spectrum the cross-peaks at 2.7/37.3 ppm [(N-4)-
CH3, R1 form] were now observed, and in the 2D LR
HETCORR spectrum the following cross-peaks via two,
three, and four bonds were detected: 1.6/81.2 ppm [(C-
2)-CH3/C-2, R1], 5.0/142.6 ppm [(N-1)-H/C-9a, R1], 4.8/
123.0 ppm [(N-1)-H/C-5a, R2], 6.8/123.0 ppm (H-9/C-5a,
R2), 2.3/169.9 ppm [(C-2)-CH3/C-2, O1], 7.3/142.6 ppm (H-
8/C-9a, R1), 1.6/142.6 ppm [(C-2)-CH3/C-9a, R1], and 1.8/
143.4 ppm [(C-2)-CH3/C-9a, R2]. The latter two cross-
peaks, which may be due to a four-bond W-arrangement
of the (C-9a)-(N-1)-(C-2)-CH3 moiety, are consistent
with the boat and inverted boat forms of 7, but not with
the chair form. These results were further considered
in connection with our molecular modeling of 7.6e
An inspection of Dreiding models shows that the
upfield shifts of the (C-2)-methyl protons from 1.79 and
1.62 ppm at 273 K to 1.66 ppm and 1.50 ppm at 192 K,
and of the (N-4)-methyl protons from 2.66 to 2.51 ppm,
and the downfield shifts of the N-1 and N-3 protons to
5.3 and 4.7 ppm, respectively, can be explained by the
proposed ring conformation in which the benzene ring
The latter in parentheses. Signal broadening.
signal at 1.7 ppm is due to the (C-2)-methyl protons of
the ring form(s). The second set of two sharper signals
in the aliphatic region, at 2.26 and 3.4 ppm, were
assigned to the (C-2)- and (N-4)-methyl protons of the
open-chain form(s), respectively. The 13C spectrum con-
firms this assumption. The peaks at 17.1 and 28.6 ppm
relate to the (C-2)-methyl carbons of the open-chain and
ring forms, respectively; the broad and weak peak at 37.9
ppm can be assigned to the (N-4)-methyl carbon, and the
weak signal at 81.2 ppm to C-2 of the ring form.
Complete assignment of these deceptively simple and
1
averaged H and 13C spectra was achieved by means of a
combined use of 2D HETCORR and 2D LR HETCORR
spectra optimized for the three-bond trans H/C configu-
ration [i.e. for J (H,C) ) 12.5 Hz]. Thus, the cross-peaks
at 28.6/1.7 ppm and 17.1/2.3 ppm in the 2D HETCORR
spectrum, and those at 170.5/2.3, 137.2/7.3, and 137.2/
2.3 ppm in the 2D LR HETCORR spectrum were used
to distinguish the two sets of peaks relating to the ring
1
and open-chain forms (see Table 3). The observed H and
13C NMR chemical shifts of the (C-2)- and (N-4)-methyl
groups reveal that the open-chain form is O1 (trans-E).

5086 J . Org. Chem., Vol. 62, No. 15, 1997
Pihlaja et al.
Ta ble 4. 1H a n d 13C NMR Ch em ica l Sh iftsa (p p m ), P op u la tion s (%), a n d Th er m od yn a m ic P a r a m eter s (k ca l m ol-1) of
Com p ou n d 7 u n d er Differ en t Con d ition sb
open (O1)e
form
trans-E
open (O2)e
form
cis-E
open (O3)d
form
trans-Z
open (O4)d
form
ring (R1)c
form
ring (R2)c
form
ring (R3)c
form
nucleus
cis-Z
(C-2)-CH3
(N-4)-CH3
(N-1)-H
1.62 (28.2)
2.70 (37.5)
5.0
1.79 (28.1)
3.46 (39.0)
4.8
1.37 (27.2)
3.30 (38.8)
2.20 (16.5)
2.17 (20.0)
3.46 (37.9f)
2.62 (26.6)
3.24 (37.9f)
2.38 (27.2)
3.10 (37.9f)
3.46 (37.9f)
4.9
(N-3)-H
4.1
4.6
-
-
-
-
population
28d
17d
26d
28d
∆Gq for
14.7 (293 K)
12.8 (253 K)
dynamic
processes
other nucleic
C-2
C-5
C-9a
(81.1)
(172.1)
(142.6)
(81.9)
(172.2)
(143.4)
(170.1)
(167.0)
(146.7)
(169.7)
(168.5)
(147.0)
a
b
d
The latter in parentheses. See the notes. c At 273 K and below. Close to the equilibrium state (after a month). e At 253 K and
below. f Splits at 233 K as follows: δ 37.2, 37.3, 35.9, 38.7, and 41.0.
up to 353 K, and on further heating up to 413 K the peaks
of O1 (ca. 57%), O3 (ca. 10%), and the ring form (ca. 33%)
were observed. Hence, the thermal (noncatalyzed) acti-
vation of the isomerization of 7 in a polar solvent shows
that the barrier height for the E/ Z (O1-O3) intercon-
version around the CdN bond is not reached until 423
K (the coalescence region was not observed), but some
increase in the O3 form population was experienced, and
thus the trans-E/ trans-Z transformation occurs via the
cis-E form. Moreover, the standard Gibbs energy differ-
ences between the open forms O1, O2, and O3 calculated
from the populations close to the equilibrium state at
room temperature (see Table 4) were found to be near
F igu r e 6. 1H NMR spectrum of 7 at near equilibrium state
(after a period of 1 month) in CDCl3 at 293 K.
zero. Thus, the dynamic process seems to be under
kinetic control, opposite to what was observed for 1-6,
where the energy of activation for the Z/ E transforma-
tion is lower (this transformation is slow at room tem-
perature and does not occur via any cis-amide form), but
rotation around the (C-2)-(C-2′) bond can affect the above
chemical shifts via the ring current effects.
Thus, at room temperature, fast rearrangements take
place including pseudorotation of the boat to the inverted
boat form, with exclusive predominance of the former.
the conformational energy differences are higher. As to
However, the time-dependent changes observed mainly
the energy of activation of E/ Z as compared to cis/ trans-
1
in the H NMR spectra of 7 within a period of more than
amide bond transformations, the situation with 7 is again
a month showed clearly that other, much slower rear-
rangements also occurred between the ring form, the ring
and open forms, and the open forms themselves. At first,
a third ring form (R3) appears, giving peaks at 1.37 ppm
opposite to that with 1-6. Below the room temperature,
1 was found only in the ring form, whereas 7 undergoes
trans-E/ cis-E transformation, with ∆Gq of ca. 12.8 kcal
mol-1 at Tc ) 253 K, i.e. lower than in the case of 1-6. A
for the (C-2)-methyl protons and at 3.30 ppm for the (N-
higher barrier for Z,E (ca. 21-24 kcal mol-1) than for
4)-methyl protons with a population of ca. 17%, whereas
the amide bond (ca. 14-17 kcal mol-1) isomerization in
the population of the (R1 + R2) forms decreases to ca.
some aroylhydrazones of methyl pyruvate has been
28%. As it concerns the open forms, up to four forms
reported and explained in terms of electronic effects.12
were observed: after the appearance of different sets of
Hence, in our compounds the replacement of a (C-2)-alkyl
peaks for O1 (27%), O2 (14%), O3 (7%), and O4 traces
by a phenyl group causes visible changes in thermody-
(see Figure 6), the amount of O2 decreases to traces, and
namic behavior, due to the different electronic and steric
properties of the substituents. An indication of such an
effect is, e.g., the about 10 ppm upfield shift of the C-5
resonance in 7-20 as compared with 1-6. As mentioned
above, MM and semiempirical MO calculations are in
progress on the open-chain isomers of these and other
model compounds and will provide a detailed insight into
at a near equilibrium state the population of O1 is ca.
26% and that of O3 is ca. 28%. The principal 1H and 13C
NMR chemical shifts of all detected forms of 7 and the
calculated conformational free energy differences be-
tween them are listed in Table 4. Thus, whereas the pair
of open-chain isomers in 3 and 4 were assigned as trans-
Z,E, O1 and O2 in 7 are trans-E and cis-E isomers [δ-
(C-2)-methyl carbon ca. 17 ppm) and the O3 and O4 forms
the problem.
While the thermodynamic behavior of the open forms
of 7 can be explained mainly in terms of changes in their
electronic properties relative to 1-6, the conformational
behavior of the ring form(s) of 7 mainly reflects the steric
effect of the phenyl substituent.
must be assigned as Z isomers [δ(C-2)-methyl carbons
ca. 27 ppm), i.e. the former signals are shifted to a higher
field due to the γ-gauche effect9,10 (already discussed
above), a consistently reliable criterion10 (see Scheme 2).
To further clarify the thermodynamics of these rear-
rangements of 7, we ran 1H and 13C NMR spectra at
higher than room temperature in DMSO-d6 solution. The
reversibility of the spectral changes was confirmed. As
in CDCl3, two sets of signals were found in the spectra
Com p ou n d s 8-20. It was of interest to inspect the
influence of substituents with different electron-with-
drawing or -releasing powers on the phenyl ring(s) of
8-20 upon the rearrangement processes described above

Conformations of Cyclic Triazepinone/Open Hydrazones
J . Org. Chem., Vol. 62, No. 15, 1997 5087
Sch em e 3
causes an increase in the R3 population at the expense
of R1; and for the counterparts 9/10 an increase in the
O2 population at the expense of O1.
Although conjugation clearly exerts some effect on the
barriers to rotation in 7-20 (Scheme 3), the complexity
of the observed ring/open-chain rearrangement processes
makes a quantitative study in terms of a simple linear
correlation: log Kx ) Fσ+ + c, where Kx ) [ring]/[chain],
found valid in our previous comparative studies of
2-aryltetrahydro-1,3-oxazines18 and 1,3-oxazolidines,19
impossible.
for 7. Such an effect may be attributed only to the
differences in their electronic properties, i.e. to resonance
effects and electrostatic interactions obeying the classical
picture of the forces that control conformational prefer-
ence.17
Exp er im en ta l Section
NMR Sp ectr a . The 1D and 2D NMR spectra were mea-
sured on a J EOL J NM GX-400 spectrometer operating at
399.78 MHz for H and 100.53 MHz for 13C, respectively, using
1
1
The H and 13C NMR spectral assignments and hence
a 5 mm 1H/13C dual probe head (90° 1H rf pulse width, 16 ms;
90° 13C rf pulse width, 8 ms; 90° decoupler rf pulse width 40
ms). All chemical shifts were referenced to the solvent [δ )
7.24 and 77.1 ppm (CDCl3), δ ) 2.49 and 39.5 ppm (DMSO-
d6), and δ ) 5.32 and 53.8 ppm (CD2Cl2) for 1H and 13C,
respectively]. The digital resolution of the 1D NMR spectra
the structure determination of 8-20 were achieved on
the basis of those of 7, as well as from the time (a few
weeks)-dependent changes in their spectra at room
temperature, assisted by 2D NMR experiments (e.g. 19).
As for 7, the initial state of the complex equilibrium
favors the ring boat (R1 and R2) form(s) in rapid
exchange and the trans-E open-chain isomer, but the
relative amounts of these forms vary essentially with the
substitution in 8-20. The dynamic processes observed
were an interconversion (i) of the open-chain trans-E (O1)
to the cis-E (O2), trans-Z (O3), and boat forms, and (ii)
of the latter to the N-3-inverted (B)′ form [i.e. (R3)] form,
giving a new O1/R1 ratio in the equilibrium state. A
temporary increase in the amount of the O2 isomer, as
for 7, was observed in most cases. Moreover, the rate of
interconversion in the above processes obviously in-
creases when going from 9 to 20. Compounds 9-16,
bearing electron-withdrawing groups (EWG), display a
higher rate of interconversion and a greater amount of
O2 and a lower amount of O3 forms at equilibrium state,
i.e. an easier amide-bond and a more difficult CdN
rotation than in 7 and 17-20. Thus, the electronic
factors controlling the conformational behavior of the
open-chain forms of 7 were strengthened in the case of
8-20. In comparison with 1-6, phenyl and/or EWG
groups at the azomethine carbon in our hydrazones lead
to a barrier-lowering effect for the cis-trans amide bond
rotation, and to a barrier-raising effect for the Z/ E-
transformation process, and both observations can be
explained in terms of electronic effects (Scheme 3).
As to the ring forms, an initial preference of R1 over
O1 (ratio ca. 2:1) and the absence of the R3 conformer
were noted for 9-16. Only for 9 and 10 two different
ring forms with δ(C-2)-methyl protons at 1.68 and 1.57
ppm were observed, just after dissolution, and were
assigned to the boat and inverted boat forms. Hence, the
barrier-raising effect of the p-NO2 group was observed
for the ring-flipping process in 9 and 10, in comparison
with 7 and 11-20. In general, a reduction in the
electron-withdrawing power and an increase in the
electron-releasing power of the substituents when going
from 9 to 20 favor the (N-3)-inverted boat (R3) form, the
amount of which increases, whereas that of R1 drastically
decreases. An essential difference in the populations of
the ring and open forms was also observed when the
effect of the chlorine substituent on the benzo part of the
benzodiazepine ring was considered. Thus, for the
counterparts 11/12 and 13/14, the chlorine substituent
1
was 0.2 and 0.7 Hz per point for H and 13C, respectively. J EOL
automatic microprograms were applied to obtain DEPT, 1H
NOE difference spectra, 2D HETCORR, and 2D LR HETCORR
1
NMR spectra. The H NOE difference spectra were acquired
with a pulse width of 8 ms, a duration of irradiation of 1000
ms, and a delay of relaxation of 8 s. The 2D experiments
normally involved a 128 × 2 K data matrix with NS ) 640
and NE ) 64. In 2D HETCORR and 2D LR HETCORR
experiments, delays of 3.33 and 2.22 ms, and 40 and 20 ms
were used to optimize the sensitivity for the response of the
1
direct and long-range H/13C coupling constants, respectively.
After zero filling in F1, a sine-bell nonshifted weighting was
used in both dimensions. The spectrometer was fitted with a
variable-temperature accessory capable of maintaining tem-
perature to within (2 K.
Syn th eses. The crystalline 1-(2′-aminobenzoyl)-1-methyl-
hydrazines and the 5′-chloro derivative were prepared accord-
ing to a literature20 method, starting from the corresponding
isatoic anhydrides with methylhydrazines. The aldehydes and
ketones used were commercial products. Melting points were
determined on an Electrothermal digital melting point ap-
paratus and are uncorrected.
A. Rea ction s of 1-(2′-Am in oben zoyl)-1-m eth ylh yd r a -
zin es w ith Alip h a tic Keton es. A mixture of 1-(2′-ami-
nobenzoyl)-1-methylhydrazine or 1-(2′-amino-5′-chlorobenzoyl)-
1-methylhydrazine (2 mmol) and 5 mL of a ketone (acetone,
3-methyl-2-butanone and 3,3-dimethyl-2-butanone were used)
was refluxed for 4-20 h. The reaction was monitored by
means of TLC. After evaporation of the solvent, products 1-5
crystallized out on ethereal treatment. The yields were 56-
67%.
B. Rea ction of 1-(2′-Am in oben zoyl)-1-m eth ylh yd r a -
zin es w ith Cycloh exa n on e. 1-(2′-Aminobenzoyl)-1-methyl-
hydrazine (1 mmol) was dissolved in 5 mL of 50% ethanol, and
0.5 g of cyclohexanone was added. The product started to
crystallize out after a few min. After standing overnight, 6
was filtered off and recrystallized from 50% EtOH in 84% yield.
C. Rea ction s of 1-(2′-Am in oben zoyl)-1-m eth ylh yd r a -
zin es w ith Acetop h en on es. A mixture of 1-(2′-aminoben-
zoyl)-1-methylhydrazine or 1-(2′-amino-5′-chlorobenzoyl)-1-
methylhydrazine (2 mmol), an acetophenone derivative (2
mmol), and p-toluenesulfonic acid (10 mg) was refluxed in
benzene (15 mL) for 5-8 h. The solvent was evaporated off,
and products 7-20 crystallized out after ethereal treatment
and were recrystallized. The yields were 70-90%.
(18) Fu¨lo¨p, F.; Pihlaja, K.; Mattinen, J .; Berna´th, G. J . Org. Chem.
1987, 52, 3821.
(19) (a) Fu¨lo¨p, F.; Berna´th, G.; Mattinen, J .; Pihlaja, K. Tetrahedron
1989, 45, 4317. (b) Fu¨lo¨p, F.; Pihlaja, K.; Neuvonen, K.; Berna´th, G.;
Argay, Gy.; Ka´lma´n, A. J . Org. Chem. 1993, 58, 1967.
(20) Fu¨lo¨p, F.; Pihlaja, K. Org. Prep. Proc. Int. 1991, 23, 377.
(17) Riddell, F. G. The Conformational Analysis of Heterocyclic
Compounds; Academic Press: 1980; p 7.

5088 J . Org. Chem., Vol. 62, No. 15, 1997
Pihlaja et al.
Ta ble 5. An a lytica l Da ta on 2,2-Disu bstitu ted -1,2,3,4-tetr a h yd r o-5H-1,3,4-ben zotr ia zep in -5-on es 1-20
anal. calcd/found, %
no.
R1
R2
X
H
mp (°C) solvent
formula MW
C
H
N
1
CH3
CH3
CH3
135-137a,b
C11H15N3O
205.26
C11H14ClN3O
239.71
C13H19N3O
233.32
C13H18ClN3O
267.76
2
3
CH3
CH3
CH3
CH3
Cl
H
152-154a
55.12
55.20
66.92
67.04
58.32
58.34
67.98
68.09
5.89
6.12
8.21
8.30
6.78
6.90
8.56
8.67
17.53
17.64
18.01
18.26
15.69
15.73
16.99
17.07
CH(CH3)2
CH(CH3)2
C(CH3)3
73-75
hexane
113-114
hexane
73-75
hexane
4
Cl
H
5
C14H21N3O
247.34
6
-(CH2)5-
H
138-140c
50% EtOH
124-125d
50% EtOH
123-125
50% EtOH
144-146
EtOH
C14H19N3O
245.32
C16H17N3O
267.33
C16H16ClN3O
301.78
C16H16N4O3
312.33
C16H15ClN4O3
346.78
C17H16F3N3O
335.33
C17H15ClF3N3O
369.78
C16H16ClN3O
301.78
C16H15Cl2N3O
336.22
C16H16BrN3O
346.24
C16H15BrClN3O
380.68
C17H19N3O
281.36
C17H18ClN3O
315.81
C17H19N3O2
297.36
C17H18ClN3O2
331.81
7
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
C6H5
C6H5
H
8
Cl
H
63.68
63.67
61.53
61.47
55.42
55.33
60.89
60.93
55.22
55.22
63.68
63.68
57.16
57.22
55.51
55.80
50.48
50.53
72.57
72.82
64.66
64.71
68.67
68.93
61.54
61.60
5.34
5.41
5.16
5.30
4.36
4.48
4.89
5.04
4.09
4.20
5.34
5.45
4.50
4.64
4.66
4.77
3.97
4.13
6.81
6.88
5.75
5.80
6.44
6.52
5.47
5.63
13.92
13.86
17.94
18.02
16.16
16.33
12.53
12.59
11.36
11.45
13.92
13.71
12.50
12.67
12.14
12.40
11.04
11.31
14.93
14.86
13.31
13.50
14.13
14.37
12.66
12.71
9
C6H4-pNO2
C6H4-pNO2
C6H4-pCF3
C6H4-pCF3
C6H4-mCl
C6H4-mCl
C6H4-pBr
10
11
12
13
14
15
16
17
18
19
20
Cl
H
180-182
EtOH
175-177a
Cl
H
169-171a
150-151
EtOH
Cl
H
141-142
EtOH
138-139a
C6H4-pBr
Cl
H
150-151a
104-106a
C6H4-pCH3
C6H4-pCH3
C6H4-pOCH3
C6H4-pOCH3
Cl
H
144-146
50% EtOH
140-142a
Cl
102-105
50% EtOH
a
b
d
Diisopropyl ether:EtOH ) 4:1. Literature5c mp 128-130 °C. c Literature5c mp 132-134 °C. Literature5c mp 132-134 °C.
Compounds 21 and 22 were prepared in ethanolic solution
at room temperature, starting from acetophenone and (o-
aminobenzoyl)hydrazine or 1-(2′-(methylamino)benzoyl)-1-
methylhydrazine, respectively. 21: mp 171-172 °C (benzene),
yield 76%; 22, mp 97-99 °C (diethyl ether), yield 56%.
Analytical data on 2,2-disubstituted-1,2,3,4-tetrahydro-5H-
1,3,4-benzotriazepin-5-ones 1-20 are shown in Table 5.
(3) In general, the dynamic processes for 1-6 are
mainly under thermodynamic control (higher conforma-
tional energy differences), while for 7-20 lower confor-
mational energy differences between the ring/open and
ring/ring forms, and also the higher ring-inversion and
(N-3)-inversion energy barriers, make the processes more
complex, and they occur under kinetic control.
Con clu sion s
(4) Larger rotational barrier heights for cis-trans-
amide bond rotation and lower heights for E/ Z CdN
bond rotation were found for alkyl-substituted hydra-
zones of 1, 3, and 4 in comparison with their phenyl(aryl)
analogues 7-20.
(5) The conjugation and hence the push-pull electronic
effects play some role in the barriers to rotation in the
open-chain hydrazones of 7-20, but, because of the
complexity of the observed rearrangements, their quan-
titative study in terms of a simple linear correlation: log
Kx ) Fσ+ + c, where Kx ) [ring]/[chain], was not possible.
(1) The studied 2,2-disubstituted-1,3,4-benzotriazepi-
nones undergo complex molecular rearrangements which
involve (i) ring to open-chain tautomerization; (ii) pseu-
dorotation of the ring forms and ring (N-3)-inversion
processes; (iii) interconversion of the open-chain forms
via cis-trans-amide bond and/or Z/ E CdN bond isomer-
izations, with different rates on the NMR time scale,
leading to the observation of average and deceptively
1
simple high-resolution H and 13C NMR spectra for most
of them in solution at room temperature.
(2) The boat form is the most stable conformation of
the ring structures of 1, 2, and 7-20, with methyl and
phenyl(aryl) groups at C-2, whereas for the sterically
congested derivatives 3 and 4, the inverted boat seems
to be the minimum energy conformer from both experi-
mental and PM3 calculations.6e
Ack n ow led gm en t. M.F.S. acknowledges a Fellow-
ship from the Turku University Foundation and in part
from the Academy of Finland.
J O962322K