Resolution and rotational barriers of quinolinone and acridone sulfenamide derivatives: Demonstration of the S-N chiral axis

Article abstract of DOI:10.1021/jo001041g

Achiral (8a) and chiral (8b) N-(2,4-dinitrobenzenesulfenyl)acridone derivatives were synthesized. Addition of the chiral solvating agent (S)- 2,2,2-trifluro-1-(anthryl)ethanol to 8a rendered the enantiotopic groups on the acridone ring diastereotopic and anisochronous, thus allowing the estimation of a lower limit for the rotational barrier about the S-N bond (18.7 kcal mol^-1) by NMR spectroscopy. 8b and the previously reported chiral sulfenamide 5 (Raban, M.; Martin, V. A.; Craine, L. J. Org. Chem. 1990, 55, 4311) were resolved on a Chiracel OD HPLC column. This constitutes the first resolution of stereostable enantiomers of a compound in which the chirality is due only to the presence of the S-N chiral axis. The rotational barriers of both compounds are nearly equal (22.7-22.8 kcal mol^-1 at 303.7 K) and are the largest determined to date for the rotation about the S-N bond in sulfenamides. The relatively high enantiomerization barrier for 8b is remarkable since the peri positions are unsubstituted.

Full text of DOI:10.1021/jo001041g

J . Org. Chem. 2000, 65, 8613-8620
8613
Resolu tion a n d Rota tion a l Ba r r ier s of Qu in olin on e a n d Acr id on e
Su lfen a m id e Der iva tives: Dem on str a tion of th e S-N Ch ir a l Axis
Merav Ben-David Blanca,†,‡ Chiyo Yamamoto,§ Yoshio Okamoto,§ Silvio E. Biali,*,‡ and
Daniel Kost*,†
Department of Chemistry, Ben Gurion University of the Negev, Beer-Sheva 84105, Israel, Department of
Organic Chemistry, The Hebrew University of J erusalem, J erusalem 91904, Israel, and Department of
Applied Chemistry, Graduate School of Engineering, Nagoya University, Chikusa-ku,
Nagoya 464-8603, J apan
kostd@bgumail.bgu.ac.il
Received J uly 11, 2000
Achiral (8a ) and chiral (8b) N-(2,4-dinitrobenzenesulfenyl)acridone derivatives were synthesized.
Addition of the chiral solvating agent (S)- 2,2,2-trifluro-1-(anthryl)ethanol to 8a rendered the
enantiotopic groups on the acridone ring diastereotopic and anisochronous, thus allowing the
estimation of a lower limit for the rotational barrier about the S-N bond (18.7 kcal mol-1) by NMR
spectroscopy. 8b and the previously reported chiral sulfenamide 5 (Raban, M.; Martin, V. A.; Craine,
L. J . Org. Chem. 1990, 55, 4311) were resolved on a Chiracel OD HPLC column. This constitutes
the first resolution of stereostable enantiomers of a compound in which the chirality is due only to
the presence of the S-N chiral axis. The rotational barriers of both compounds are nearly equal
(22.7-22.8 kcal mol-1 at 303.7 K) and are the largest determined to date for the rotation about the
S-N bond in sulfenamides. The relatively high enantiomerization barrier for 8b is remarkable
since the peri positions are unsubstituted.
In tr od u ction
depending on whether in the transition state the S-R
bond adopts a syn-periplanar arrangement with the R2
Sulfenamides of the general formula 1 adopt a confor-
mation in which the R1SN and R2NR3 planes are on
average nearly perpendicular, and hence, the S-N bond
constitutes a chiral axis as long as R2 * R3.1 The objective
of the present work is to directly demonstrate the
chirality of the S-N axis by optical resolution and
isolation of stereostable enantiomers, owing their optical
activity to the presence of the S-N chiral axis.
or the R3 substituent (Chart 1).
The magnitude of the rotational barrier in sulfen-
amides depends on a combination of steric and electronic
effects. Electron withdrawing substituents on the sulfur
atom and bulky substituents on the nitrogen are known
to increase the torsional barrier.1 In most cases examined
to date the rotational barriers were not sufficiently high
to allow for the resolution of enantiomers, and hence were
measured using only NMR methods.4 We have recently
described the first enantiomeric resolution of a system
(2) which owes its chirality to the sulfenamide chiral axis
and the determination of its optical activity.5 However,
the enantiomers of 2 proved to be stereostable only at
0 °C. The nitrogen atom in sulfenamide 2 is nearly planar
due to conjugation with the aromatic system, and the
enantiomerization process consists almost entirely of a
T process. The two bulky substituents flanking the S-N
bond hinder the T process, thus raising the rotational
barrier.6 The enantiomerization barrier of 2 (determined
When the nitrogen is pyramidal and its two substitu-
ents are different (R2 * R3) four stereoisomeric forms (two
enantiomeric pairs) are possible. Enantiomerization of
a given form requires both torsion (T) around the S-N
bond and inversion (I) of the nitrogen atom (Figure 1).2
If the nitrogen atom is nearly planar or inverts rapidly,
enantiomerization is dominated by torsion about the S-N
bond.3a,b Two diastereomeric rotational pathways exist,
(3) (a) Symmetry arguments indicate that, precluding accidental
degeneracy, in a chiral sulfenamide (with R2 * R3) the nitrogen and
the three atoms connected to it should not lie exactly on the same
plane. The symmetry operations necessary to force such arrangement
are either a symmetry plane (σ) passing through the four atoms or a
C2 axis collinear with the R-N or N-S bonds. Since the molecule
belongs to the C1 point group, neither symmetry operation can be
present and a planar arrangement cannot be enforced. For a discussion
on symmetry arguments see: Mislow, K. Introduction to Stereochem-
istry, W. A. Benjamin, NY, 1965; p. 8. (b) A similar stereochemical
† Ben Gurion University of the Negev.
‡ The Hebrew University of J erusalem.
§ Nagoya University.
(1) For reviews on the stereochemistry of sulfenamides see: (a) Kost,
D.; Raban, M. in The Chemistry of Sulphenic Acids and their Deriva-
tives, Patai, S., Ed., Wiley, Chichester, 1990, ch 2. (b) Raban, M.; Kost,
D. in Acyclic Organonitrogen Stereodynamics, Lambert, J . B.; Takeuchi,
Y., Ed., VCH: New York, 1992; p. 57. (c) Raban, M.; Kost, D.
Tetrahedron 1984, 40, 3345.
analysis applies also for
a time scale under which the nitrogen
inversion process is fast relative to the torsion around the S-N bond.
(4) A barrier equal to or higher than 23 kcal mol-1 is usually
sufficient to separate two interconverting species at 298 K (Kessler,
H. Angew. Chem., Int. Ed. Eng. 1970, 9, 219).
(2) For a discussion of dynamic processes in a similar model system
see: Kost, D.; Aviram, K.; Raban, M. J . Org. Chem. 1989, 54, 4903.
(5) Ben-David Blanca, M.; Maimon, E.; Kost, D. Angew. Chem., Int.
Ed. Engl. 1997, 36, 2216.
10.1021/jo001041g CCC: $19.00 © 2000 American Chemical Society
Published on Web 11/21/2000

8614 J . Org. Chem., Vol. 65, No. 25, 2000
Ben-David Blanca et al.
and the S-Ar group in the rotational transition state is
expected to yield higher rotational barriers. Notably,
evidence on related compounds suggests that such sys-
tems might possess substantial rotational barriers even
in the absence of peri substituents (i.e., R ) H in 4). Raban
and co-workers have shown that although the 2-quino-
linone sulfenamide 6 has an enantiomerization barrier
of only 16.9 kcal mol-1, in the 4-quinolinone derivative 5
this barrier is higher than 22 kcal mol-1.9 The different
barrier heights of 5 and 6 were interpreted as indicating
that in both cases the lower energy rotational pathway
involves passage of the dinitrophenyl group near the Me
(or CdO) substituent (an “exo” passage). It was concluded
that the passage near the peri hydrogen introduces
severe steric interactions leading to a barrier in excess
of 22 kcal mol-1.9 In this paper, we report the prepara-
tion, resolution of enantiomers, and determination of the
enantiomerization barrier of a chiral acridone sulfen-
amide derivative lacking any peri substituent, and of
sulfenamide 5.
F igu r e 1. Interconversion graph of the four isomeric forms
(two enantiomeric pairs) of a sulfenamide derivative with a
pyramidal nitrogen atom. Pairs of forms located at opposite
corners of the graph represent enantiomers. Nitrogen inversion
and torsion about the S-N bond are denoted by the letters “I”
and “T”, respectively.
Ch a r t 1
from the coalescence of the methyl signals in the 1H NMR
spectrum) was ∆G‡ ) 21.0 kcal mol-1 (C6D5NO2, 417 K).
After 24 min at 0 °C the resolved sample had completely
racemized. In the present study we sought to demon-
strate the chirality generated by the sulfenamide chiral
axis directly by preparing and resolving stereostable
enantiomers. Chirality would then be manifested in the
persistent optical activity of the enantiomers, by contrast
to previously investigated sulfenamides, in which both
chirality and rotational barriers could only be observed
and determined by NMR methods.
Isolation of enantiomers which are stereostable at the
laboratory time scale demands a higher barrier of rota-
tion around the S-N bond. Two alternative routes for
further increasing the enantiomerization barrier come to
mind. The first route involves retaining the central
scaffold of the amino part of the sulfenamide (the
carbazole unit) and increasing the bulk of the substitu-
ents at the peri positions flanking the sulfenamide bond.7
A different approach involves creating a more crowded
environment about the S-N bond by changing the
relative disposition of the peri substituents. This can be
achieved, for example, by the formal replacement of the
central five membered ring of the carbazole moiety of 3
by a six membered ring (e.g., 4). The smaller exocyclic
R1 angles in 4 should result in a closer disposition of the
peri substituents to the sulfenamide bond.8 The increased
repulsive steric interaction between the peri substituents
Resu lts a n d Discu ssion
Str u ctu r a l Con sid er a tion s a n d Syn th esis. Two
acridone sulfenamide derivatives lacking peri substitu-
ents were chosen as targets: 8a and 8b. In both cases
the acridone nitrogen is expected to be nearly planar due
to conjugation of its lone pair with the aromatic system.
The isopropyl group in 8b was introduced to serve two
purposes: (i) to desymmetrize the acridone moiety ren-
dering the molecule chiral and (ii) to serve as a prochiral
probe capable of “sensing” the chirality of the molecule
and allowing NMR monitoring of the enantiomerization
process.
(6) The 2,4-dinitrophenyl substituent was chosen since the rotational
barrier of sulfenamides N-substituted by this group is higher than
when substituted by the more crowded 2,4,6-trinitrophenyl group.
See: Raban, M.; Yamamoto, G. J . Am. Chem. Soc. 1977, 99, 4160; J .
Am. Chem. Soc. 1979, 101, 5890.
(7) Preliminary attempts to synthesize sulfenamide derivatives by
reaction of the anions of crowded carbazoles with 2,4-dinitrobenzene-
sulfenyl chloride proved unsuccessful. Apparently, the increased steric
crowding hinders the formation of the sulfenamide bond.
(8) ZINDO calculations of the 2-isopropyl derivatives of 4 and 3
(Ar ) 2,4-(NO2)2C6H3) indicate that in the first compound the R1 angle
should be 7° smaller than in the second. (Maimon, E. M.Sc. thesis,
Ben Gurion University, 1993).
(9) Raban, M.; Martin, V. A.; Craine, L. J . Org. Chem. 1990, 55,
4311.

Rotational Barriers of Acridone Sulfenamide Derivatives
J . Org. Chem., Vol. 65, No. 25, 2000 8615
F igu r e 2. 1H NMR (400 MHz, C2D2Cl4) of the achiral sulfenamide 8a .
Sch em e 1
by ca. 0.6 ppm in the sulfenamides 8a and 8b as a result
of their proximity to the dinitrophenyl ring.
The isopropyl methyl groups of 8b are accidentally
isochronous in CDCl3, but become anisochronous in
acetone-d6 or DMF-d7 solutions although they are sepa-
rated by only 1.3 and 1.9 Hz, respectively, at 400 MHz
(Figure 3). The rather similar shift of the signals is due
to their relatively large distance from the sulfenamide
chiral axis. The anisochrony of the isopropyl methyl
groups is evidence that at room temperature 8b exists
in a “frozen” chiral conformation (relative to the NMR
time scale).
Rota tion a l Ba r r ier of Su lfen a m id e 8a . Sulfenamide
8a is achiral and a 180° rotation around the S-N bond
results in homomerization which mutually exchanges the
two phenylene rings. These rings are enantiotopic under
slow rotation (on the NMR time scale) around the
sulfenamide bond, and homotopic under fast rotation.
Since enantiotopic groups are isochronous under achiral
conditions, the rotational process is “invisible” and cannot
be followed by NMR. Enantiotopic groups can be ren-
dered anisochronous in a chiral nonracemic medium, and
therefore the homomerization process of 8a could be
followed in such a medium. Addition of the chiral
solvating agent (CSA) (S)-2,2,2-trifluoro-1-(anthryl)etha-
nol (10)13 to a C2D2Cl4 solution of 8a resulted in noticeable
splitting of the acridone protons in the 400 MHz 1H NMR
spectrum, while the dinitrophenyl protons remained
unsplit as expected. The largest splitting (∆δ ) 18.7 Hz)
was observed for the H5 protons (Figure 4). Due to partial
Isopropyl acridone (7b) was prepared according to
Scheme 1.10 Acridones 7a and 7b were deprotonated with
KH, and the resulting anions were treated with 2,4-
dinitrobenzenesulfenyl chloride yielding the yellow sul-
fenamides 8a and 8b.11 Sulfenamide 5 was prepared by
a minor modification of the literature procedure.9,12
NMR Sp ectr a . Sulfenamide 8a displays in the 1H
NMR spectrum (400 MHz, CDCl3, rt) seven aromatic
signals (Figure 2). Assignment of the aromatic signals
of 8a and 8b was carried out on the basis of their coupling
pattern and in the case of 8b, by COSY and NOESY
experiments. Notably, the peri protons vicinal to the
nitrogen (H7 in Figure 2), which in the parent acridones
7a and 7b resonate at ca. 7.3 ppm, are shifted downfield
(10) (a) Agranat, I.; Tapuhi, Y. J . Am. Chem. Soc. 1978, 100, 5604.
(b) Allen, C. F. H.; McKee, G. H. W. Org. Synt. Coll. Vol. II, p 15.
(11) We were unable to prepare by this method the dinitroben-
zenesulfenyl derivative of peri-substituted acridones, probably due to
steric hindrance effects of the alkyl substituent on the nitrogen atom.
(12) The “c” and “d” arrows of Scheme 1 in ref 9 should point to the
structures 13 and 12, respectively, and not as depicted.
(13) Pirkle, W. H.; Beare, S. D. J . Am. Chem. Soc. 1969, 91, 5150.

8616 J . Org. Chem., Vol. 65, No. 25, 2000
Ben-David Blanca et al.
F igu r e 3. 1H NMR (400 MHz, acetone-d6) spectrum of 8b (expansion of the isopropyl methyl region). The two partially overlapping
doublets correspond to the diastereotopic isopropyl methyl groups.
overlap of the triplets of each proton, their combined
patterns form a quartet.
were conducted. The experiments were initially con-
ducted in o-dichlorobenzene-d4 solution at 446 K. At high
temperatures the compound substantially decomposed to
its starting material 11. To minimize the exposure of 5
to high temperatures, the NMR probe was preheated to
446 K using an ethylene glycol sample standard. The
ethylene glycol was replaced by a sample of 5 and the
NMR spectrum determined after thermal equilibrium
had been achieved (10-15 min). Under these conditions,
ca. 50% of the compound decomposed but the unreacted
5 displayed two signals for the benzylic diastereotopic
protons, indicating that the enantiomerization process
was slow on the NMR time scale. From the chemical shift
difference (19.9 Hz) between the proton signals and their
mutual coupling constant (J ) 15.3 Hz), a lower limit of
49 s-1 can be estimated for the exchange rate at 446 K
by the Kurland equation.14a From this rate constant a
lower limit of 23.0 kcal mol-1 was estimated for the
enantiomerization barrier.
Upon raising the temperature of 8a in the presence of
10 the two H5 signals merged (Figure 4). This process
can be attributed either to a fast rotation (on the NMR
time scale) around the S-N bond (a homomerization) or
to a weaker association between the chiral additive and
8a . Since the ∆δ value between the two H5 protons
diminished upon raising the temperature and none of the
line shape changes characteristic of a coalescence process
were observed (line broadening, formation of a plateau
between the coalescing signals), it may be concluded that
the observed process is due to a weaker association. On
the basis of the chemical shift difference at room tem-
perature (∆δ ) 6 Hz) and the highest temperature at
which separate signals could be distinguished for the two
H5 protons (350 K), a lower limit of ∆Gq > 18.7 kcal mol-1
could be estimated for the rotational barrier of 8a .
High -Tem p er a tu r e NMR Stu d ies of 5 a n d 8b. The
enantiomerization barriers of 5 and 8b can be determined
from the rate of exchange of the diastereotopic protons
(the methylene protons of the benzyl group in 5 and the
isopropyl methyl groups in 8b) at a given temperature.
Raban and co-workers reported that at 300 MHz in 4:1
o-dichlorobenzene-d4/toluene-d8 solution the diastereotopic
methylene proton signals of 5 do not undergo any line
broadening or coalescence up to 423 K. On the basis of
these data, a lower limit of 22 kcal mol-1 was estimated
for the enantiomerization barrier of 5.9 To obtain a better
estimate of the lower limit of its enantiomerization
barrier, additional high-temperature NMR studies of 5
The high-temperature NMR studies of 8b were con-
ducted in DMF-d7. Increasing the sample temperature
resulted in broadening of the partially overlapping iso-
propyl methyl signals and coalescence. Partial decompo-
sition of the sample was observed during the high-
temperature studies. From the chemical shift difference
of the isopropyl methyls (∆δ ) 1.9 Hz) the rate of
exchange (k ) 4.2 s-1) at the coalescence temperature
(394 K) corresponding to a barrier of ∆Gq ) 22.1 kcal
mol-1 was estimated by the Gutowsky-Holm equation.14b
Due to the proximity of the two partially overlapping
doublets, the error introduced by the use of the Gu-
towsky-Holm equation can be relatively large.15 Full line
shape analysis of the spectra at 380, 384, 389, and 394
(14) (a) Kurland, R. J .; Rubin, M. B.; Wise, W. B. J . Chem. Phys.
1964, 40, 2426. (b) Gutwosky, H. S.; Holm, C. H. J . Chem. Phys. 1956,
25, 1228.
(15) For a critical examination of the validity of the coalescence
methods see: Kost, D.; Carlson, E. H.; Raban, M. J . Chem. Soc., Chem.
Commun. 1971, 656; Kost, D.; Zeichner, A. Tetrahedron Lett. 1974,
4533.

Rotational Barriers of Acridone Sulfenamide Derivatives
J . Org. Chem., Vol. 65, No. 25, 2000 8617
F igu r e 4. 1H NMR (400 MHz, C2D2Cl4) spectra of a 1:15 mixture of 8a and the chiral solvating agent 10 at different temperatures.
The strong signals in the middle of the spectrum correspond to the chiral solvating agent.
K was conducted with the gNMR program.16 The simula-
tion provided rate constants of 2, 2.5, 3.0, and 4.0 s-1
,
respectively, which correspond to a barrier of ∆Gq
)
22.0 ( 0.1 kcal mol-1
.
Resolu tion a n d En a n tiom er iza tion Ba r r ier s. Ini-
tial attempts to resolve the sulfenamide 8b by enantio-
selective HPLC were conducted at room temperature
using a Chiracel OD (12) column (eluent: hexane/EtOH
8:2). Although the two enantiomers were resolved, the
chromatogram displayed a plateau between the peaks of
the enantiomers, indicating that some enantiomerization
took place during the time scale of the resolution.17
(16) gNMR v4.1.0. Cherwell Scientific Publishing: Oxford, UK.

8618 J . Org. Chem., Vol. 65, No. 25, 2000
Ben-David Blanca et al.
F igu r e 5. Chromatogram of the enantioseparation of 8b by HPLC on Chiralcel OD at 10 °C. The opposite signs of the CD peaks
of the resolved sample indicate that the species separated indeed correspond to enantiomeric forms.
Ta ble 1. En a n tiom er iza tion Ra te Con sta n ts (k) a n d
Kin etic Activa tion P a r a m eter s for th e S-N Rota tion of
5 a n d 8b
It was previously observed that the racemization
barrier determined by following the decay of the CD
spectra of 2 was lower than the barrier measured by
NMR spectroscopy.5 This could indicate that under the
CD irradiation a photochemically induced racemization
pathway is operative.19 Solutions of resolved samples of
5 and 8b were kept in the dark at 293.2 and 288.2 K,
respectively, and the progress of the enantiomerization
at those temperatures was followed by enantioselective
HPLC. The rate constants obtained for 5 and 8b (6.66 ×
10-5 and 3.70 × 10-5 s-1, respectively) were practically
equal to those obtained at the same temperatures by
following the process by CD (6.63 × 10-5 and 3.59 × 10-5
s-1, Table 1), indicating that there is no photochemical
contribution to the racemization process in either of the
two compounds.
T (K)
k (s-1
)
∆Gq (kcal mol-1
)
5a
293.6
298.6
303.7
308.7
313.7
288.6
293.7
303.7
308.7
313.7
6.63 × 10-5
1.34 × 10-4
2.62 × 10-4
4.78 × 10-4
9.38 × 10-4
3.59 × 10-5
6.60 × 10-5
2.88 × 10-4
5.97 × 10-4
1.14 × 10-3
22.8
22.8
22.8
22.8
22.7
22.8
22.8
22.7
22.7
22.6
8bb
a
b
∆Hq ) 23.3 kcal mol-1, ∆Sq ) 1.80 cal K-1 mol-1
.
∆Hq ) 24.6
kcal mol-1, ∆Sq ) 6.19 cal K-1 mol-1
.
Baseline separation of the enantiomers was obtained at
10 °C (Figure 5). The racemization of resolved samples
of 5 and 8b was followed by CD at different temperatures
(Figure 6).18 Excellent first order rate plots were obtained
from the CD data. The kinetic data are collected in Table
1. The Gibbs free energies of activation of both com-
pounds are nearly equal (22.8-22.7 kcal mol-1 at 303.7
K) and are ca. 2 kcal mol-1 larger than the values
determined for the enantiomerization of 2 by the coales-
cence method (∆Gq ) 21.0 kcal mol-1). The increase in
the rotational barrier can be ascribed to the increase in
the central ring size from five membered in 2 to six
membered in 5 and 8b. The relatively high enantiomer-
ization barrier of 8b is remarkable since the peri posi-
tions are unsubstituted. Since the rotational barriers of
8a and 8b should be similar, it must be concluded that
the spectral changes observed upon heating of a sample
of 8a in the presence of 10 (see above) must be due to
weaker association at the elevated temperature.
Exo vs P er i Rota tion a l Tr a n sition Sta tes in th e
Ch ir a l Su lfen a m id e 5. Two diastereomeric transition
states are possible for 5, involving the passage of the
dinitrophenyl group near the peri hydrogen (a peri
transition state) or near the methyl group (an exo
transition state). The large difference between the enan-
tiomerization barriers of 5 and 6 was previously inter-
preted as indicating that in both compounds the lowest
energy enantiomerization pathway (threshold pathway)
involves the exo transition state.9 However, although the
large difference in barriers obviously rules out a common
peri transition state for both compounds (the barriers in
that case should be similar), an additional possibility is
that the threshold enantiomerization pathways for the
two compounds are different, namely exo for 6 and peri
for 5.
The two diastereomeric transition states for the enan-
tiomerization of 8b are expected to possess essentially
similar energies. In those transition states the repulsive
steric interactions involving a peri hydrogen and the
rotating ring should be similar to those present in the
peri transition state of 5. The nearly equal enantiomer-
(17) For a discussion of the peak coalescence phenomena in enan-
tioselective chromatography see: Schurig, V. Chirality 1998, 10, 140.
See also: Gasparrini, F.; Lunazzi, L.; Mazzanti, A.; Pierini, M.;
Pietrusiewicz, K. M.; Villani, C. J . Am. Chem. Soc. 2000, 122, 4776.
(18) The slope of a plot of ln(ee) vs time affords the racemization
rate which is twice the enantiomerization rate (krac ) 2 kenant).
(19) See for example: Okamoto, Y.; Honda, S.; Yuki, H.; Nakamura,
H.; Iitaka, Y.; Nozoe, T. Chem. Lett. 1984, 1149.

Rotational Barriers of Acridone Sulfenamide Derivatives
J . Org. Chem., Vol. 65, No. 25, 2000 8619
in the absence of bulky peri substituents, suggesting a
significant effect of the exocyclic angle (R1) size on steric
congestion. The enantiomerization barriers of 5 and 8b
represent the highest barriers determined to date for the
sulfenamide bond.
Exp er im en ta l Section
Ch r om a togr a p h ic Resolu tion s. Resolution of the sul-
fenamides 5 and 8b were conducted using a Chiracel OD
column (25 × 0.46 cm i.d.) and a UV/CD J ASCO CD-1595, 254
nm detector. The resolutions were conducted at 0 °C (5) or 10
°C (8b) using 20:80 (5) or 60:40 (8b) hexane/EtOH mixtures
as eluents and a flow rate of 0.5 mL/min. The determination
of the enantiomeric excess by HPLC was conducted at 0 °C.
NMR Exp er im en ts. Temperatures were measured using
ethylene glycol spectra, and are assumed accurate within (1
K.20 To minimize the sample decomposition in the high-
temperature studies, the NMR probe was heated to the desired
temperature with an ethylene glycol sample, which was
replaced by the sulfenamide sample once the temperature was
achieved. NMR spectra were recorded after 15 min to ensure
thermal equilibration of the sample.
N-(2,4-Din itr oben zen esu lfen yl)a cr id on e (8a ). Dry THF
(25 mL) was added via a syringe to a flask charged with
acridone (0.195 g, 1 mmol) and potassium hydride (0.06 g, 1.5
mmol) under an argon atmosphere. After the evolution of
hydrogen ceased, the flask was cooled to -78 °C and a solution
of 0.282 g (1.2 mmol) of 2,4-dinitrobenzenesulfenyl chloride
in 15 mL of dry THF was added dropwise. The flask was
allowed to warm to room temperature over 18 h, the THF was
evaporated and the reaction mixture partitioned between
chloroform and water. The organic layer was dried (anhydrous
MgSO4) and evaporated and the residue chromatographed
(silica gel, eluent: CHCl3) to yield 0.18 g (46%) pure 8a as a
yellow solid: mp 220 °C dec; 1H NMR (CDCl3, 300.13 MHz) δ
9.27 (d, J ) 2.3 Hz, 1 H), 8.60 (dd, J ) 7.9 Hz, 1.4 Hz, 2H),
8.21 (dd, J ) 8.9 Hz, 2.4 Hz, 1H), 7.97 (d, J ) 8.5 Hz, 2H),
7.69 (dt, J ) 7.9 Hz, 1.7 Hz, 2H), 7.44 (t, J ) 7.8 Hz, 2 H),
6.98 (d, J ) 9.0 Hz, 1 H) ppm; 13C NMR (DMSO-d6, 100.1 MHz)
δ 176.96, 146.13, 145.50, 143.46, 143.36, 134.78, 128.95,
126.87, 125.08, 123.56, 123.51, 121.18, 118.02 ppm; CI MS m/z
(rel intensity) 393.9 (MH+).
N-(4-Isop r op ylp h en yl)a n th r a n ilic Acid (9). A mixture
of 22.7 mL (0.166 mol) of p-isopropyl aniline (cumidine), 4.1 g
(0.026 mol) of o-chlorobenzoic acid, 4.1 g (0.03 mol) of anhy-
drous K2CO3, and 0.1 g of copper(I) oxide was refluxed for 2 h.
The excess aniline was removed by steam distillation. 2 g of
charcoal and 20 mL of water were added to the brown residue,
and the mixture was boiled for 15 min and filtered. A mixture
of 3 mL of concentrated HCl and 6 mL of water was added to
the stirred filtrate. The purple-brown acid which precipitated
after cooling was filtered and purified by dissolving in a hot
solution of 2.5 g Na2CO3 in 100 mL water, adding 2.5 g of
charcoal, boiling for 10 min, filtering and acidifying. The yield
of the beige acid was 4.2 g, 63%: mp 170-174 °C (lit.10a mp
172-174 °C); 1H NMR (CDCl3, 400.13 MHz) δ 9.27 (br s, 1 H,
NH), 8.05 (dd, J ) 8.1, 1.4 Hz, 1 H), 7.35 (dt, J ) 7.7 Hz, 1.5
Hz, 1 H), 7.24-7.16 (m, 6 H), 6.73 (dt, J ) 7.0 Hz, 1.1 Hz, 1
H), 2.88-2.95 (q, J ) 7.0 Hz, 1 H), 1.28 (d, J ) 6.9 Hz, 6 H)
ppm; 13C NMR (CDCl3, 100.1 MHz) δ 173.19, 149.56, 145.16,
137.95, 135.22, 132.61, 127.43, 123.71, 116.79, 113.99, 109.97,
33.72, 24.15 ppm.
2-Isop r op yla cr id on e (7b). A mixture of 3.2 g (12.5 mmol)
of 9 and 10 mL of concentrated sulfuric acid was heated on a
boiling water bath for 4 h. The mixture was poured slowly into
100 mL of boiling water with stirring and after boiling for 5
additional min filtered by suction. The yellow-brown paste was
boiled for 10 min with a solution of 3 g of Na2CO3 in 40 mL of
water, and the crude product was filtered and washed with
water. The greenish solid was recrystallized from hot DMSO
and washed with CHCl3 to give 0.6 g (65%) of 7b: mp 265 °C;
F igu r e 6. Decay of the CD spectra of resolved samples of 5
(top, in 2:8 hexane/ethanol at 303.7 K) and 8b (bottom, in 6:4
hexane/ethanol at 308.7 K).
ization barriers determined for 5 and 8b suggest (disre-
garding ground-state effects) that the rotating rings
experience rather similar steric environments in the
rotational transition states of both compounds. It may
be concluded that the peri pathway is operative in 5, and
that either the peri and exo pathways are similar in
energy, or that the enantiomerization proceeds largely
via the peri transition state.
Con clu sion s
Enantiomeric resolution of 5 and 8b, which owe their
chirality to the presence of the S-N chiral axis, was
accomplished on an enantioselective HPLC column. Thus,
the chirality generated by the S-N chiral axis, in the
absence of any other stereogenic unit, has been directly
demonstrated. First-order rate constants for the enan-
tiomerization process were determined. Sulfenamide
derivatives of acridone possess substantially higher
rotational barriers than their carbazole analogues, even
(20) Van Geet, A. L. Anal. Chem. 1968, 40, 2227.

8620 J . Org. Chem., Vol. 65, No. 25, 2000
Ben-David Blanca et al.
1H NMR (CDCl3, 400.13 MHz) δ 8.49 (d, J ) 7.6 Hz, 1 H),
8.33 (d, J ) 1.8 Hz, 1 H), 8.24 (s, 1 H, NH), 7.63 (dt,
J ) 8.4, 1.5 Hz, 1 H), 7.56 (dd, J ) 8.5, 2.0 Hz, 1 H), 7.31-
7.24 (m, 3 H), 3.05 (q, J ) 6.9 Hz, 1 H), 1.31 (d, J ) 6.9 Hz, 6
H) ppm; 13C NMR (CDCl3, 100.1 MHz) δ 176.57, 140.93,
140.68, 139.19, 133.10, 132.52, 125.92, 122.08, 120.63, 120.24,
120.19, 117.35, 117.15, 32.83, 23.78 ppm.
121.94, 117.09, 116.89, 33.50, 23.87 ppm; CI MS, m/z 435.9
(MH+). Anal. Calcd: C, 60.68; H, 3.93; N, 9.65. Found: C,
60.38; H, 4.17; N, 9.86.
3-Ben zyl-2-m eth yl-4(1H)-qu in olin on e (11). The com-
pound was prepared by a minor modification of the literature
procedure.9 A mixture of 8.8 g (0.04 mol) of ethyl 2-benzyl-
acetoacetate, 3.72 g (0.04 mol) of distilled aniline, 12 g of
anhydrous CaSO4 (Drierite), 0.08 mL of glacial acetic acid, and
12 mL of absolute ethanol was refluxed for 5 h. The reaction
mixture was filtered, the solids were washed with ethanol, and
the combined filtrates were evaporated. Distillation of impuri-
ties took place when the mixture was heated to 140 °C at 5
mmHg using a sand bath. Cyclization occurred when heating
the sand bath to 280 °C for 20 min. After cooling, the mixture
solidified and the crude quinolinone was treated with charcoal
in boiling ethanol, filtered and recrystallized from absolute
ethanol to yield 4.7 g (48%) 11, mp 284-6 °C (lit.9 mp 290.5-
294.5 °C).
N-(2,4-dinitrobenzenesulfenyl)-2-isopropylacridone (8b).
Under an argon atmosphere, 25 mL of dry THF were added
via a syringe to a flask charged with 0.2 g (0.84 mmol) of 7b
and 0.05 g (1.25 mmol) of KH. After evolution of hydrogen had
ceased, the flask was cooled to -78 °C and a solution of 0.233
g (1 mmol) of 2,4-dinitrobenzenesulfenyl chloride in 15 mL of
dry THF was added dropwise. The flask was allowed to warm
to room temperature over 18 h, the THF was evaporated, and
the reaction mixture partitioned between chloroform and
water. The organic layer was dried over anhydrous MgSO4,
and the solvent was evaporated. The sulfenamide was purified
by chromatography (silica gel, eluent: CHCl3) to yield 0.3 g
(55%) of pure 8b as a bright orange solid: mp 218-220 °C
Reaction of the potassium salt of 11 with 2,4-dinitroben-
zenesulfenyl chloride according to the literature procedure9
afforded 5.
1
(from acetonitrile); H NMR (CDCl3, 400.13 MHz) δ 9.26 (d,
J ) 2.3 Hz, 1 H), 8.60 (dd, J ) 8.0 Hz, 1.5 Hz, 1 H), 8.44 (d,
J ) 2.2 Hz, 1 H), 8.21 (dd, J ) 9.0 Hz, 2.4 Hz, 1 H), 7.95 (d,
J ) 8.6 Hz, 1 H), 7.88 (d, J ) 8.9 Hz, 1 H), 7.67 (dt, J ) 8.5
Hz, 1.4 Hz, 1 H), 7.57 (dd, J ) 8.9 Hz, 2.2 Hz, 1 H), 7.41 (t,
J ) 7.3 Hz, 1 H), 6.97 (d, J ) 9.0 Hz, 1 H), 3.07 (q, J ) 7.0 Hz,
1 H), 1.32 (d, J ) 6.9 Hz, 6 H) ppm; 13C NMR (CDCl3, 100.1
MHz) δ 177.87, 147.00, 146.00, 144.93, 143.27, 141.52, 134.62,
133.74, 128.73, 128.32, 125.41, 125.17, 124.27, 124.19, 123.83,
Ack n ow led gm en t. We thank Dr. Roy Hoffman for
assistance with the NMR experiments.
Su p p or tin g In for m a tion Ava ila ble: 1H NMR spectrum
of 8a in DMSO-d6. This material is available free of charge
via the Internet at http://pubs.acs.org.
J O001041G