Synthesis and stereochemistry of N-phenyl-N-9-triptycylhydroxylamine derivatives and related compounds

Article abstract of DOI:10.1246/bcsj.77.2273

During the course of attempts at synthesis of N-phenyl-N-9- triptycylhydroxylamine (1), we found that the reaction of N-9- triptycylhydroxylamine (5) with benzenediazonium-2-carboxylate afforded 3-(2-carboxyphenyl)-1-(9-triptycyl)-triazene 1-oxide (6), th

Full text of DOI:10.1246/bcsj.77.2273

ꢀ 2004 The Chemical Society of Japan
Bull. Chem. Soc. Jpn., 77, 2273–2281 (2004) 2273
Synthesis and Stereochemistry of N-Phenyl-N-9-triptycylhydroxylamine
Derivatives and Related Compounds
ꢀ
Chiharu Agawa, Keiko Otsuka, Mao Minoura, Yasuhiro Mazaki, and Gaku Yamamoto
Department of Chemistry, School of Science, Kitasato University, Kitasato, Sagamihara, Kanagawa 228-8555
Received July 30, 2004; E-mail: gyama@kitasato-u.ac.jp
During the course of attempts at synthesis of N-phenyl-N-9-triptycylhydroxylamine (1), we found that the reaction
of N-9-triptycylhydroxylamine (5) with benzenediazonium-2-carboxylate afforded 3-(2-carboxyphenyl)-1-(9-triptycyl)-
triazene 1-oxide (6), the structure of which was confirmed by X-ray crystallography. Oxidation of N-(9-triptycyl)aniline
(4) with mCPBA gave phenyl 9-triptycyl nitroxide (7) and N-(9-triptycyl)-1,4-benzoquinonimine (8) in small yields. Re-
duction of 7 with phenylhydrazine gave 1, and O-ethylation of 1 gave O-ethyl-N-phenyl-N-9-triptycylhydroxylamine (2).
Results of X-ray crystallography for compounds 2, 4, 7, and 8 and results of dynamic NMR studies for 1, 2, and 8 are
presented. In compound 8, Tp–N rotation and C=N topomerization took place on the NMR timescale at high temper-
atures above 60 ꢁC, and the energy barriers to both processes were obtained. In compounds 1 and 2, dynamic NMR be-
ꢁ
havior was observed at very low temperatures below ca. ꢂ50 C, and the energy barriers to the respective processes
could not be determined except for those related to ‘‘Ph-passing’’.
We have hitherto reported on the static and dynamic stereo-
chemistry of various triptycene derivatives carrying a nitrogen
HOOC
O
atom at the bridgehead, such as N,N-dialkyl-9-triptycyl-
amines,1,2 N-(9-triptycyl)anilines,3,4
N-(9-triptycyl)acet-
H
N
N
N
O
O
amides,5–7 and N-(9-triptycyl)acetanilide.8 We have recently
been interested in N-9-triptycylhydroxylamines and have re-
ported on the molecular structures in crystal and the dynamic
behavior in solution of those carrying alkyl group(s) on the ni-
trogen or oxygen atom or on both.9–12 We naturally planned to
study the corresponding N-aryl derivatives of the hydroxyl-
amine, because significantly different behavior has often been
observed between the N-alkyl and N-aryl derivatives of the
above-mentioned compounds. We have thus made various at-
tempts at syntheses of N-phenyl-N-9-triptycylhydroxylamine
(1) and its O-alkyl derivatives (Chart 1). After several unsuc-
cessful attempts, compound 1 and its O-ethyl derivative 2
could be obtained, and the dynamic NMR studies on these
compounds will be described in this article. We found some
interesting chemistry during the course of these studies, which
will also be reported.
N
N
6
7
8
Chart 2.
Since it had been established that benzyne reacted with 9-
triptycylamine to give N-(9-triptycyl)aniline (4)4 by N–H in-
sertion of the benzyne, thermal decomposition of benzenedi-
azonium-2-carboxylate, a well-known precursor of benzyne, in
the presence of N-9-triptycylhydroxylamine (5) was examined.
We expected the addition of benzyne to the NHOH moiety, but
the product was found to be 3-(2-carboxyphenyl)-1-(9-tripty-
cyl)triazene 1-oxide (6), a direct addition product of benzene-
diazonium-2-carboxylate with 5 as described below (Chart 2).
Then we attempted the oxidation of N-(9-triptycyl)aniline
(4) to give N-phenyl-N-9-triptycylhydroxylamine (1). Oxida-
tion with mCPBA was thus examined, but the reaction afforded
only a small amount (ca. 7%) of phenyl 9-triptycyl nitroxide
(7), together with N-(9-triptycyl)-1,4-benzoquinonimine (8),
also in a similar small yield. An N-alkyl-1,4-benzoquinon-
imine is rather a rare type of molecule and the molecular struc-
ture and dynamic behavior of 8 was investigated in detail. Re-
duction of the nitroxide 7 gave the hydroxylamine 1 and ethyl-
ation of 1 gave the O-ethyl derivative 2. Dynamic NMR stud-
ies of these compounds are described below.
At the outset, we attempted reactions of organometallic
compounds with nitroso compounds, because such reactions
have been known to be the typical preparation methods of hy-
droxylamines. Reaction of phenyllithium with 9-nitrosotripty-
cene resulted in the formation of O-phenyl-N-9-triptycylhy-
droxylamine (3), instead of the desired N-phenyl derivative
1, presumably by way of the single electron transfer process,
which has been reported previously.9
R
R'
1: R = C6H5, R' = OH
2: R = C6H5, R' = OC2H5
3: R = H, R' = OC6H5
4: R = C6H5, R' = H
5: R = H, R' = OH
N
Results and Discussion
Reaction of Benzenediazonium-2-carboxylate with N-9-
Triptycylhydroxylamine. Benzenediazonium-2-carboxylate
Chart 1.
Published on the web December 10, 2004; DOI 10.1246/bcsj.77.2273

2274 Bull. Chem. Soc. Jpn., 77, No. 12 (2004)
N-9-Triptycylhydroxylamines
although amines are generally far more basic than the corre-
sponding hydroxylamines: the pKb values have been reported
to be 8.03, 8.04, 8.80, 9.40, 9.25, and 10.35 for NH2OH,
MeNHOH, Me2NOH, NH2OMe, MeNHOMe, and Me2NOMe,
respectively,14 and 4.75, 3.38, 3.23, and 4.20 for NH3, MeNH2,
Me2NH, and Me3N, respectively.15 A literature survey re-
vealed that the above-mentioned type of reaction is not unpre-
cedented: N-phenylhydroxylamine, for example, reacted with
diazotized anthranilic acid to give 3-(2-carboxyphenyl)-1-
phenyltriazene 1-oxide,16 the structure of which was confirmed
by X-ray crystallography.17
O2C
OH
NH
CO2
N
Tp
N
N
N
N
9
Tp-NH-OH
OH
CO2H
O
N
H
N
CO2H
N
N
Tp
Tp
N
6
10
Scheme 1.
Oxidation of N-(9-Triptycyl)aniline with m-Chloroper-
benzoic Acid. Oxidation of secondary amines to the corre-
sponding hydroxylamines or nitroxides has sporadically been
reported; many of the researchers used m-chloroperbenzoic
acid (mCPBA) as the oxidant. Therefore, oxidation of 4 with
mCPBA was examined, which proved to give phenyl 9-tripty-
cyl nitroxide (7) together with N-(9-triptycyl)-1,4-benzoquino-
nimine (8), both in small amounts.
The nitroxide 7 has been reported as the spin-trap product of
phenyl radical with 9-nitrosotriptycene,18 where only the ESR
data were given without further characterization. The ESR data
found in our compound 7 agreed well with those reported in
Ref. 18. X-ray crystallographic analysis of 7 was performed
in order to confirm the structure: the molecular structure is
shown in Fig. 2(a) and the representative bond lengths and an-
gles are given in Table 1. The phenyl ring is nearly coplanar
with the Tp–N bond and almost bisects the notch between
two benzene rings of the Tp group. The nitrogen is slightly
pyramidalized: the sum of the bond angles around the nitrogen
is 358.0ꢁ. This is contrasted with the complete planarity of the
nitroxide moieties found recently in an alkyl aryl nitroxide19
and a dialkyl nitroxide,20 and can be ascribed to the steric ef-
fect of the bulky Tp moiety.
Reduction of the nitroxide 7 with phenylhydrazine gave N-
phenyl-N-9-triptycylhydroxylamine (1) almost quantitatively
as shown by NMR, the isolation yield being 83%. Compound
1 was rather unstable and easily oxidized to 7 in the air. Depro-
tonation of 1 with potassium t-butoxide followed by treatment
with ethyl iodide gave O-ethyl-N-phenyl-N-9-triptycylhydrox-
ylamine (2). X-ray crystallographic and dynamic NMR studies
on this compound will be described in a later section.
The molecular structure of N-(9-triptycyl)-1,4-benzoquinon-
imine (8), the other oxidation product of compound 4, is shown
in Fig. 2(b) together with that of 4 in Fig. 2(c) for comparison.
The representative bond lengths and angles are also compiled
in Table 1. The quinone ring in 8 or the phenyl ring in 4 is co-
planar with the Tp–N bond and almost bisects the notch of the
Tp moiety.
O3
C6
C7
O2
C5
N3
N2
O1
C4
N1
C2
C1
C3
Fig. 1. The ORTEP drawing with 50% probability ellip-
soids of compound 6.
(9),13 prepared by aprotic diazotization of anthranilic acid, was
heated in boiling chloroform in the presence of N-9-triptycyl-
hydroxylamine (5) to afford a white solid mass in a high yield,
which was finally identified to be 3-(2-carboxyphenyl)-1-(9-
triptycyl)triazene 1-oxide (6). Mass spectral analysis suggested
the product to be a 1:1 adduct of 9 with 5. Although IR and
NMR spectra could not definitely differentiate 6 from the tau-
tomeric hydroxytriazene 10 (Scheme 1), X-ray crystallograph-
ic analysis confirmed the structure of 6.
The molecular structure of 6 is shown in Fig. 1. Selected
lengths and angles are given in the Experimental Section.
The triazene oxide moiety is almost coplanar with the o-
caboxyphenyl group (N2–N3–C5–C6: ꢂ168:5ꢁ) and nearly
eclipses one of the benzene rings of the Tp moiety (N2–N1–
C1–C2: 20.0ꢁ). The NH moiety intramolecularly hydrogen
bonds both with the N-oxide oxygen and the carbonyl oxygen
ꢀ
(O1 N3: 2.473 A; O2 N3: 2.653 A). This will be coincident
with the appearance of the NH proton at a very low field (ꢀ
13.5) in the H NMR spectrum. Two molecules of 6 associate
ꢀ
ꢃꢃꢃ
ꢃꢃꢃ
1
to form a hydrogen-bonded dimer by way of the carboxy
groups.
A plausible mechanism of formation of 6 is given in
Scheme 1.
At first, the facile reaction of benzenediazonium-2-carbox-
ylate (9) with the hydroxylamine 5 before decomposing to
benzyne seemed surprising, because the corresponding amine,
TpNH2, does not react with 9 but with benzyne formed by de-
composition of 9 and gives 4 in a good yield.4 This indicates
that TpNHOH has far stronger nucleophilicity than TpNH2,
Dynamic NMR Studies of the Quinonimine 8. Com-
pound 8 shows an interesting stereochemical behavior in solu-
tion in the sense that two types of dynamic processes are pos-
sible: rotation about the Tp–N bond and topomerization of the
C=N double bond. Actually these two processes were ob-
served by NMR spectroscopy in different temperature ranges.
At room temperature, both processes are slow on the NMR
timescale: one of the Tp-benzene ring is different from the oth-
er two and two edges of the quinone ring are mutually non-
equivalent. As the temperature of the sample in 1,1,2,2-tetra-

C. Agawa et al.
Bull. Chem. Soc. Jpn., 77, No. 12 (2004) 2275
(a)
(b)
(c)
C5
C6
C5
C6
C6
C5
N
N
N
C1
C1
C1
C4
C4
C4
C3
C3
C3
C4
C4
C4
C2
C2
C2
Fig. 2. The ORTEP drawings with 50% probability ellipsoids of (a) compound 7, (b) compound 8, and (c) compound 4.
ꢁ
ꢀ
Table 1. Selected Bond Lengths (A) and Angles ( ) for
Compounds 7, 8, and 4aÞ
Temp / °C
Compound
7
8
4
126
N–C5
N–C1
C1–C2
C1–C3
C1–C4
1.434(2)
1.488(2)
1.553(2)
1.552(2)
1.554(2)
1.298(2)
1.459(2)
1.552(2)
1.551(2)
1.546(2)
1.398(3)
1.444(3)
1.540(3)
1.541(3)
1.531(3)
111
95
C1–N–C5
O–N–C5
O–N–C1
N–C1–C2
N–C1–C3
N–C1–C4
122.8(1)
117.4(1)
117.8(1)
109.7(1)
114.0(1)
116.2(2)
123.6(2)
—
—
114.8(2)
117.2(2)
109.2(1)
126.4(2)
—
—
114.8(2)
114.6(2)
112.0(2)
C1–N–C5–C6
C5–N–C1–C2
C5–N–C1–C3
C5–N–C1–C4
7.4(2)
ꢂ72:6ð1Þ
50.2(1)
ꢂ0:8ð2Þ
ꢂ67:6ð2Þ
57.8(2)
ꢂ8:7ð4Þ
ꢂ59:1ð3Þ
65.5(3)
ꢂ176:7ð2Þ
80
59
168.8(1)
175.0(2)
a) The atom numbering is shown in Fig. 2.
8/13
C
2/7/15
5/16
3/6/14
chloroethane-d2 was raised, the Tp-aromatic signals began
broadening at ca. 60 ꢁC and coalesced above 90 ꢁC, while
the quinone ring signals began broadening only above 110
ꢁC. The aromatic region spectra of 8 at several temperatures
are shown in Fig. 3.
A
1
D
4
B
23
7.5
6.5
δ
8.0
7.0
The lineshape change of the Tp-aromatic signals is ascribed
to the rotation of the Tp–N bond, and lineshape analysis of the
signals due to 4-, 5-, and 16-H at six temperature points in the
range of 80–100 ꢁC gave the rate constant at each temperature.
In addition, saturation transfer experiments were performed by
observing the intensity change of the 13-H signal (ꢀ 8.12) upon
irradiation of the 1,8-H signal (ꢀ 6.88) to afford rate constants
at five temperature points in the range of 39–59 ꢁC. Least
Fig. 3. 1H NMR spectra of compound 8 in the aromatic
region at several temperatures in 1,1,2,2-tetrachloroeth-
ane-d2.
squares analysis of the Eyring plot including the whole tem-
perature range gave the following kinetic parameters: ꢁHz ¼
74:4 ꢄ 1:3 kJ molꢂ1, ꢁSz ¼ ꢂ14:4 ꢄ 3:9 J molꢂ1 Kꢂ1, ꢁGz
(80 ꢁC) = 79.5 kJ molꢂ1. This is the highest energy barrier

2276 Bull. Chem. Soc. Jpn., 77, No. 12 (2004)
N-9-Triptycylhydroxylamines
R
O
tBu
tBu
Et
Et
HN
HN
R
N
N
PhCH2
Ph
12
13
ground state
transition state
Chart 3.
Scheme 2.
(a)
O
O
HB
HA
HD
HC
H
H
H
H
N
N
C2
N
Tp
Tp
C6
C5
Scheme 3.
C4
C3
to rotation of the Tp–N bond hitherto known. The ꢁGz values
for the Tp–N rotation so far reported are, for example, 45.6
kJ molꢂ1 (ꢂ59 ꢁC) in N-(9-triptycyl)aniline (4)4 and 70.2
kJ molꢂ1 (26 ꢁC) in N-2-biphenylyl-9-triptycylamine (11).4
These compounds at the ground state adopt a conformation
in which the aryl or quinone ring is almost coplanar with the
Tp–N bond (Scheme 2), as revealed by the molecular struc-
tures in crystal (see Fig. 2). At the transition state for the
Tp–N rotation, the aryl or quinone plane should significantly
twist in order to relieve the steric repulsion against the Tp moi-
ety. Compound 4 can easily adopt such a conformation as
shown in Scheme 2 at the transition state (R ¼ H), while com-
pound 11 (R ¼ C6H5) is subject to significant steric congestion
even in a twisted conformation at the transition state, and this
causes the barrier in 11 to be far higher than that in 4. Com-
pound 8 finds it even harder to adopt a twisted conformation
similar to the one shown in Scheme 2 at the transition state be-
cause the nitrogen constitutes a double bond, and thus the bar-
rier is even higher.
O
C7
(b)
C6
C4
C5
C2
N
C3
O
C7
Fig. 4. The ORTEP drawings with 50% probability ellip-
soids of (a) molecule A and (b) molecule B of com-
pound 2.
The second dynamic process of 8 observed at higher tem-
peratures is the topomerization of the C=N double bond
(Scheme 3). Because coalescence of the quinone-ring protons
could not be expected below the boiling point of the solvent
used (145 ꢁC), lineshape analysis was abandoned, and the rate
constants for this process were obtained by saturation transfer
experiments at four temperature points in the range of 110–125
ꢁC. Here the change in the signal intensity of HD (ꢀ 7.81,
see Fig. 3 and Scheme 3) was followed upon irradiation of
HB (ꢀ 6.57). Least squares analysis of the Eyring plot gave
7
O
5
N
6
4
3
1
2
Chart 4.
the following kinetic parameters: ꢁHz ¼ 86:2 ꢄ 4:5 kJ molꢂ1
ꢁSz ¼ ꢂ22:3 ꢄ 11:5 J molꢂ1 Kꢂ1, ꢁGz (126 ꢁC) = 95.1
kJ molꢂ1
,
of compound 2 contained two kinds of molecules in a crystal
lattice; these are referred to as molecules A and B. Perspective
drawings of these molecules are shown in Fig. 4, and selected
bond lengths, bond angles, and torsion angles are compiled in
Table 2.
These two molecules have opposite chirality for the nitrogen
and are significantly different in their conformations and geo-
metries. The structure of molecule A is rather similar to the
structures of the N,O-dialkyl-N-9-triptycylhydroxylamines
hitherto reported.12 The nitrogen has an almost tetrahedral ge-
ometry (the sum of the angles around nitrogen is 332.3ꢁ), and
the C1–N–O–C7 dihedral angle (149.3ꢁ) is characteristic of an
N-9-triptycylhydroxylamine as discussed in detail previously
(see Chart 4 for the atomic numbering).12 This angle results
.
A lot of studies have been reported on the isomerization or
topomerization of C=N double bonds.21 The isomerization/
topomerization can take place either by rotation around the
C=N bond or planar inversion at the nitrogen atom, and theo-
retical studies have indicated that the latter generally has a
lower energy barrier than the former. Therefore, planar inver-
sion is tentatively assigned to the observed process. The ꢁGz
values have been reported to be 102.5 and 92.0 kJ molꢂ1 for
the C=N bond topomerization in compounds 1222 and 13,23 re-
spectively (Chart 3).
X-ray Crystallography of Compound 2. A single crystal

C. Agawa et al.
Bull. Chem. Soc. Jpn., 77, No. 12 (2004) 2277
ꢁ
ꢀ
Table 2. Selected Bond Lengths (A) and Angles ( ) for the
Two Molecules of Compound 2aÞ
a
N
O
N
O
Molecule
N–O
A
B
R
R
1.465(2)
1.466(2)
1.552(2)
1.546(2)
1.554(2)
1.452(2)
1.439(2)
1.404(2)
1.469(2)
1.558(2)
1.560(2)
1.552(2)
1.407(2)
1.451(2)
b
b
N–C1
C1–C2
C1–C3
C1–C4
N–C5
O–C7
O
O
N
R
R
N
a
a
R
O
R
O
C1–N–O
C1–N–C5
C5–N–O
Sum
105.2(1)
117.6(1)
109.5(1)
332.3
117.5(1)
124.1(1)
112.2(1)
353.8
b
N
N
N–C1–C2
N–C1–C3
N–C1–C4
N–O–C7
112.2(1)
110.3(1)
118.3(1)
111.3(1)
115.7(2)
116.0(2)
110.0(1)
114.1(1)
a: "Ph-passing"; b: "O-passing"
Scheme 4.
or ‘‘O-passing’’, respectively4). Each passing process is accom-
panied by inversion of the nitrogen and partial rotation of the
N–O bond. In the ‘‘Ph-passing’’, the phenyl group necessarily
rotates by ca. 180ꢁ and thus the two edges interchange (so-
called ‘‘Gear Rotation’’ or GR).4 The phenyl group can also ro-
tate inside the notch made of the two flanking benzene rings,
independently of the ‘‘Ph passing’’ process (so-called ‘‘Isolated
Rotation’’ or IR).4
C1–N–O–C7
C5–N–O–C7
O–N–C1–C2
O–N–C1–C3
O–N–C1–C4
C5–N–C1–C2
C5–N–C1–C3
C5–N–C1–C4
C1–N–C5–C6
O–N–C5–C6
149.3(1)
ꢂ83:4ð2Þ
176.7(1)
ꢂ70:2ð1Þ
50.8(2)
77.6(2)
ꢂ75:8ð2Þ
ꢂ99:2ð2Þ
25.3(2)
139.5(1)
50.8(2)
175.3(2)
ꢂ70:5ð2Þ
12.8(2)
54.5(2)
167.6(1)
ꢂ71:4ð2Þ
25.3(2)
ꢂ94:7ð2Þ
At room temperature, completely conformationally aver-
aged NMR spectra were obtained for both compounds; the
three benzene rings of the Tp moiety were mutually equivalent
and the N-phenyl proton signals appeared as an A2B2C spin
system. The methylene protons of the ethyl group in 2 gave
a sharp quartet signal. Many of the signals broadened and de-
coalesced with the decrease of the temperature, but completely
re-sharpened spectra were not obtained eveꢁn at the lowest tem-
164.2(1)
a) The atom numbering is shown in Chart 4 and Fig. 4.
from the steric repulsion between the O-methylene group and
the nearby benzene ring of the Tp moiety, which distorts the
angle from the ‘‘ideal’’ value of 120ꢁ. The phenyl plane is tilt-
ed from the C1–N–C5 plane by ca. 25ꢁ because of the steric
interaction with the O-methylene group, and is almost perpen-
dicular to the O–N–C5 plane. In molecule B, on the contrary,
the nitrogen is highly planarized (the sum of the angles around
nitrogen is 353.8ꢁ), and the N–O and N–C5 bonds are not as
long as those of molecule A. The phenyl plane is tilted from
the C1–N–C5 and O–N–C5 planes to similar degrees (ca.
15ꢁ). One may reasonably infer that the nitrogen lone-pair con-
jugates with the phenyl group and thus diminishes the electro-
static repulsion with the oxygen lone-pairs, resulting in the
shortening of both the N–C5 and N–O bonds. The C1–N–O–
C7 dihedral angle is 77.6ꢁ and thus the O–C7 bond almost
bisects the C1–N–C5 angle. In other words, the O–C7 bond
is antiperiplanar to the nitrogen lone-pair, and this conforma-
tion corresponds to the second stable one in hydroxylamines
(see Scheme 2 of Ref. 12). This is the first example of such
a conformation found in crystal.
Dynamic NMR Study of Compounds 1 and 2. The plau-
sible mechanistic pathway of conformer interconversion in N-
phenyl-N-9-triptycylhydroxylamines, 1 and 2, is schematically
shown in Scheme 4 by the analogy of the N-alkyl derivatives
discussed previously.12 The molecule is assumed to reside in a
conformation similar to molecule A in the crystal of 2. Thus,
one conformer transforms to its enantiomeric conformer by
the passing of either the phenyl group or the OR group over
a benzene ring of the Tp moiety (referred to as ‘‘Ph-passing’’
ꢁ
peratures examined (ꢂ100 C and ꢂ115 C for 1 and 2, re-
spectively). Spectra of the aromatic region at several tempera-
tures are shown in Figs. 5 and 6 for 1 and 2, respectively.
At the lowest temperatures, some of the signals could be as-
signed by decoupling and qualitative saturation transfer experi-
ments, and the assignments are given in Figs. 5 and 6. In com-
pound 1, the edge-exchange of the phenyl group, i.e., the rota-
tion of the N–Ph bond, was slow on the NMR timescale at
ꢂ100 ꢁC, and the two ortho-protons (20- and 60-H) were sepa-
rately observed: the 20-H signal appeared at a considerably
high field of ꢀ 6.32 because of the ring-current effect of the
flanking benzene rings of the Tp moiety (Fig. 5). The chemical
shift is similar to that of the 20-H signal in compound 4 (ꢀ 6.15
ꢁ
at ꢂ87 C),4 indicating that the phenyl plane is almost copla-
nar with the C9–N–C10 plane (Chart 5). From the chemical
shift difference between the 20-H and 60-H signals (ca. 350
Hz), the rate constant for the N–Ph rotation was roughly esti-
mated to be ca. 780 sꢂ1 at the plausible coalescence tempera-
ꢁ
ture (ca. ꢂ60 C), corresponding to ꢁGz of ca. 40 kJ molꢂ1
.
The N–Ph rotation takes place by two processes, GR (Ph-pass-
ing) and IR, as mentioned above, and thus the rate constant of
ca. 780 sꢂ1 is the sum of the rate constants for these two proc-
esses. This means that ꢁGz of both GR and IR should not be
less than 40 kJ molꢂ1
.

2278 Bull. Chem. Soc. Jpn., 77, No. 12 (2004)
N-9-Triptycylhydroxylamines
3/6/15
3/6/15
2/7/14
1/8/13
4/5/16
2'/6'
3'/5'
4/5/16
3'/4'/5'
Temp / °C
−8
Temp / °C
−33
2'/6'
1/8/13
2/7/14
4'
−32
−48
−59
−73
−87
−63
−71
−80
4'
4/5/16
−101
−115
6'
4/5/16
4'
2'
−100
5'
7/14
2'
6'
δ
8.0
7.5
7.0
6.5
Fig. 5. 1H NMR spectra of compound 1 in the aromatic
region at several temperatures in dichloromethane-d2.
The numberings are shown in Chart 5.
δ
8.0
7.5
7.0
6.5
Fig. 6. 1H NMR spectra of compound 2 in the aromatic
region at several temperatures in dichloromethane-d2.
The numberings are shown in Chart 5.
5'
6'
4'
1'
RO
3'
14
N
9
In compound 2, the signals due to the phenyl group were
separately identified at ꢂ115 ꢁC except for 30-H, indicating
the slow rotation of the N–Ph bond (Fig. 6). The 20-H signal
appeared at ꢀ 7.17, a considerably lower field than the corre-
sponding signal in compound 1, probably because of the tilting
of the phenyl plane out of the C9–N–C10 plane. From the
chemical shift difference between the 20-H and 60-H signals
(ca. 195 Hz), the rate constant for the N–Ph rotation was
roughly estimated to be ca. 430 sꢂ1 at a plausible coalescence
temperature (ca. ꢂ90 ꢁC), corresponding to ꢁGz of ca. 35
kJ molꢂ1.ꢁThe signals due to the Tp moiety were still broad
at ꢂ115 C. The 4/5/16-H signals appeared as 2:1 doublets
as did those in 1, indicating that either Ph-passing or O-passing
is slow but the other is fast at this temperature. A rough
estimate of the rate constant of ca. 15 sꢂ1 and ꢁGz of ca. 36
kJ molꢂ1 at ꢂ108 ꢁC was obtained for the site exchange.
2'
8
13
1
4
2
3
7
6
10
5
Chart 5.
As for the Tp-proton signals of 1, those due to 4/5/16-H ap-
peared as two doublets with the intensity ratio of 2:1 at ꢂ100
ꢁC, and the other signals were obscure because of the signal
broadening and overlapping (Fig. 5). As can be inferred from
Scheme 4, three benzene rings of the Tp moiety are equivalent
when both Ph-passing and O-passing are fast, and are mutually
nonequivalent when both processes are slow, while two are
equivalent and one is different when either one process is slow
and the other is fast. The appearance of the 4/5/16-H signals
ꢁ
ꢁ
at ꢂ100 C corresponds to the last situation. The lineshape of
The methylene signal remained a broad quartet at ꢂ115 C,
ꢁ
the 4/5/16-H signal at ꢂ80 C gave a rough estimate of the
in agreement with the reasoning that either Ph-passing or O-
passing is still fast at this temperature. By similar reasoning
as in 1, one can infer that ꢁGz of the Ph-passing process in
rate constant of ca. 25 sꢂ1 and thus ꢁGz of ca. 41 kJ molꢂ1
for the site-exchange, which corresponds to the slower of the
two processes. This means that both Ph-passing and O-passing
have ꢁGz of ca. 41 kJ molꢂ1 or lower. From these results and
those given in the preceding paragraph, ꢁGz of the Ph-passing
process is concluded to lie between 40 and 41 kJ molꢂ1, while
the ꢁGz values of the IR and O-passing processes are not de-
termined.
2 lies between 35 and 36 kJ molꢂ1
.
The results of the DNMR studies for 1 and 2 are given in
Table 3, together with the data for the related compounds pre-
viously reported. Both the R-passing and O-passing barriers in
1 and 2 are considerably lower than the respective counterparts
in N-methyl derivatives 14 and 15. The lower barrier to R-

C. Agawa et al.
Bull. Chem. Soc. Jpn., 77, No. 12 (2004) 2279
Table 3. The Results of DNMR Studies for 1, 2, and Related CompoundsaÞ
ꢁGz/kJ molꢂ1
Compd
R
R1
Ref
R-passing
O-passing
IR
1
2
4
14
15
Ph
Ph
Ph
Me
Me
OH
OEt
H
OH
OEt
40–41
35–36
45.6
52.6
57.7
ꢅ41
ꢅ36
—
51.0
46.5
ꢆ40
ꢆ35
44.8
—
This work
This work
4
12
12
—
a) Measured in dichloromethane-d2.
R
CH3O
CH3O
H2C
N
16 :R = CH3
17: R = C6H5
OR
Chart 6.
Chart 7.
ꢁ
passing in the N-phenyl compounds will be ascribed mainly to
stabilization of the transition state by dynamic gearing (an
analogy with that shown in Scheme 2). A similar situation
was observed in 9-ethyl-1,4-dimethoxytriptycene (16) and its
9-benzyl counterpart 17; the barrier to rotation of the 9-sub-
stituent was 61.1 kJ molꢂ1 for 16 and 49.8 kJ molꢂ1 for 17
(Chart 6).24
It is notable that the GR (i.e., R-passing) barrier in 2 is lower
than that in 1, while that in 15 is higher than that in 14. The
latter has been discussed in terms of the steric interaction of
the OR1 moiety with the Tp group at the transition state, result-
ing in the higher barrier in 15 than in 14.12 The former would
be ascribed mainly to the larger destabilization of the ground
state in 2 than in 1 due to steric repulsion between the phenyl
and the O-methylene groups.
The fact that the GR barrier is higher in 4 than in 1 or in 2
will again be mainly ascribed to ground state destabilization in
1 and 2, though some other factors will contribute: the Ph
plane in 4 is coplanar with the C1–N–C5 plane (see Fig. 2
for numbering), while that in 2 is considerably (ca. 25ꢁ) tilted
as shown in Fig. 4(a).
The fact that the O-passing barriers in 1 and 2 are lower than
in 14 and 15 probably results from the lower nitrogen inversion
barrier in the N-phenyl compounds: the plausible transition
state for O-passing in 1 and 2 will be stabilized by the phenyl
group because of the conjugation with the nitrogen lone-pair
(Chart 7).
glycol sample and are reliable to ꢄ1 C. The ESR spectrum of 7
was obtained on a JEOL JES-TE200 spectrometer.
3-(2-Carboxyphenyl)-1-(9-triptycyl)triazene 1-Oxide (6).
To a boiling solution of 1.01 g (3.54 mmol) of N-9-triptycylhy-
droxylamine (5)25 in 50 mL of chloroform was added during the
course of 30 min a suspension in chloroform (30 mL) of benzene-
diazonium-2-carboxylate (9)13 prepared from 725 mg (5.29 mmol)
of anthranilic acid, and the mixture was then heated under reflux
for 30 min. The white precipitates formed were collected by filtra-
tion to afford 1.356 g (3.13 mmol, 88%) of 6, mp 208–210 ꢁC
(from chloroform). Found: C, 74.54; H, 4.60; N, 9.64%. Calcd
for C27H19N3O3: C, 74.81; H, 4.42; N, 9.69%. MS m=z 433
(Mþ). IR (KBr) 3400–2400 (br), 1678 (s) cmꢂ1. 1H NMR (CDCl3)
ꢀ 5.427 (1H, s, 10-H), 7.04–7.16 (7H, m, 2/7/14-H, 3/6/15-H, 40-
H), 7.442 (3H, m, 4/5/16-H), 7.615 (1H, dd, J ¼ 8:6, 7.2, 1.6 Hz,
50-H), 7.758 (1H, dd, J ¼ 8:5, 0.9 Hz, 60-H), 7.891 (3H, m, 1/8/
13-H), 8.219 (1H, dd, J ¼ 8:0, 1.4 Hz, 30-H), 13.49 (1H, s, NH).
The carboxy proton was not detected in CDCl3 but was found at ꢀ
8.31 in DMSO-d6. 13C NMR (CDCl3) ꢀ 53.59 (1C, t), 84.28 (1C,
q), 113.47 (1C, t), 114.37 (1C, t), 121.03 (1C, t), 122.97 (3C, t),
123.05 (3C, t), 124.94 (3C, t), 125.77 (3C, t), 131.70 (1C, q),
134.09 (1C, q), 141.71 (3C, q), 142.15 (1C, q), 143.84 (3C, q),
169.15 (1C, q).
mCPBA Oxidation of N-(9-Triptycyl)aniline. To an ice-
chilled solution of 500 mg (1.45 mmol) of N-(9-triptycyl)aniline
(4)4 in 40 mL of dichloromethane under argon was added 500
mg (2.90 mmol) of m-chloroperbenzoic acid in 60 mL of dichloro-
methane. The mixture was stirred for 1 h at 0 ꢁC and then for 1 h
at room temperature. The solution was washed with aq. NaHCO3,
dried over MgSO4 and evaporated. Column chromatography on
silica gel with dichloromethane–hexane gave 32 mg (6%) of phen-
yl 9-triptycyl nitroxide (7) and 40 mg (7%) of N-(9-triptycyl)-1,4-
benzoquinonimine (8) together with 205 mg (41%) of unreacted 4.
Phenyl 9-Triptycyl Nitroxide (7). Reddish-brown solids, mp
236–244 ꢁC (dec) (from dichloromethane–hexane). Found: C
86.70; H, 4.99; N, 3.91%. Calcd for C26H18NO: C, 86.64; H,
5.03; N, 3.89%. MS m=zꢁ 360 (Mþ, 60%), 345 (85%), 252
(100%). ESR (CH2Cl2, 22 C) g ¼ 2:0074, a ¼ 11:1 (t), 2.1 (q),
and 0.9 (t) G (lit.18 a ¼ 11:1, 2.2, and 0.9 G in benzene).
N-(9-Triptycyl)-1,4-benzoquinonimine (8). Colorless crys-
tals, mp 254–260 ꢁC (dec) (from dichloromethane–hexane).
Experimental
General. Melting points are not corrected. Mass spectra were
obtained on a Hitachi M-2500 spectrometer in the EI mode. IR
spectra were obtained on a Jasco FT/IR-610 spectrometer. 1H
and 13C NMR spectra were obtained on a Bruker ARX-300 spec-
1
trometer operating at 300.1 MHz for H and 75.4 MHz for 13C,
respectively, at 22–24 ꢁC, unless otherwise stated. Chemical shifts
are referenced with internal tetramethylsilane (ꢀH ¼ 0:00) or
CDCl3 (ꢀC ¼ 77:00). Letters p, s, t, and q given along with the
13C chemical shifts denote primary, secondary, tertiary, and qua-
ternary, respectively. In variable-temperature experiments, tem-
peratures were calibrated using a methanol sample or an ethylene

2280 Bull. Chem. Soc. Jpn., 77, No. 12 (2004)
N-9-Triptycylhydroxylamines
Found: C, 86.80; H, 4.81; N, 3.81%. Calcd for C26H17NO: C,
1
diethyl ether was added a suspension of 42 mg (0.374 mmol) of
KOtBu in 15 mL of diethyl ether, and the mixture was stirred
for 0.5 h at 0 ꢁC. To the mixture was added 0.09 mL (1.24 mmol)
of ethyl iodide and the mixture was stirred for 1 h at 0 ꢁC, and then
overnight at room temperature. The mixture was washed with wa-
ter and dried over MgSO4. Column chromatography on silica gel
with dichloromethane–hexane (1:3) as the eluent afforded 32 mg
(0.082 mmol, 82%) of 2, mp 266–272 ꢁC (dec) (from dichloro-
methane–hexane). Found: C, 86.19; H, 6.14; N, 3.77%. Calcd
for C28H23NO: C, 86.34; H, 5.95; N, 3.60%. 1H NMR (CDCl3)
ꢀ 1.259 (3H, t, J ¼ 7:1 Hz, CH3), 4.441 (2H, q, J ¼ 7:1 Hz,
CH2), 5.324 (1H, s, 10-H), 6.891 (3H, td, J ¼ 7:6, 1.5 Hz, 2/7/
14-H), 6.968 (3H, td, J ¼ 7:3, 1.3 Hz, 3/6/15-H), 7.18 (1H, m,
p-H), 7.30 (2H, m, m-H), 7.372 (3H, dd, J ¼ 7:2, 1.3 Hz, 4/5/
16-H), 7.41 (2H, m, o-H), 7.543 (3H, dm, J ¼ 7:6 Hz, 1/8/13-
H). 13C NMR (CDCl3) ꢀ 13.62 (1C, p), 54.62 (1C, t), 68.46
(1C, s), 77.42 (1C, q), 123.05 (3C, t), 124.19 (3C, t), 124.92
(3C, t), 125.28 (1C, t), 125.46 (2C, t), 125.79 (3C, t), 128.21
(2C, t), 144.21 (3C, q), 145.22 (3C, q), 146.70 (1C, q).
X-ray Crystallography. Crystals of compounds 2, 4, 6, 7, and
8 were grown from dichloromethane–hexane. The crystal data and
the parameters for data collection, structure determination, and re-
finement are summarized in Table 4. Diffraction data were collect-
ed on a Rigaku AFC7R or a Rigaku/MSC Mercury CCD diffrac-
tometer and calculations were performed using the SHELXL97
program.26 The structures were solved by direct methods followed
by full-matrix least-squares refinement with all non-hydrogen
atoms anisotropic and all hydrogen atoms isotropic. Reflection
86.88; H, 4.77; N, 3.90%. H NMR (CDCl3) ꢀ 5.492 (1H, s, 10-
H), 6.219 (1H, dd, J ¼ 10:4, 2.3 Hz, 20-H), 6.573 (1H, dd,
J ¼ 10:4, 2.7 Hz, 30-H), 6.848 (2H, d, J ¼ 7:7 Hz, 8/13-H),
6.852 (1H, dd, J ¼ 10:4, 2.3 Hz, 60-H), 6.967 (2H, td, J ¼ 7:5,
1.3 Hz, 7/14-H), 6.992 (1H, td, J ¼ 7:0, 1.3 Hz, 3-H), 7.080
(2H, td, J ¼ 7:3, 1.2 Hz, 6/15-H), 7.087 (1H, td, J ¼ 7:5, 1.3
Hz, 2-H), 7.376 (1H, dd, J ¼ 7:3, 1.0 Hz, 4-H), 7.498 (2H, dd,
J ¼ 7:3, 1.1 Hz, 5/16-H), 7.777 (1H, J ¼ 10:1, 2.7 Hz, 50-H),
8.069 (1H, dd, J ¼ 7:5, 1.2 Hz, 1-H). 13C NMR (CDCl3) ꢀ
53.75 (1C, t, 10-C), 74.07 (1C, q, 9-C), 121.81 (2C, t), 122.30
(1C, t), 123.20 (1C, t), 124.01 (2C, t), 124.58 (2C, t), 125.09
(1C, t), 125.19 (1C, t), 126.01 (2C, t), 129.58 (1C, t), 131.76
(1C, t), 132.67 (1C, t), 143.69 (1C, t), 143.69 (2C, q), 144.28
(2C, q), 144.57 (1C, q), 147.34 (1C, q), 162.44 (1C, q, C=N),
187.71 (1C, q, C=O).
N-Phenyl-N-9-triptycylhydroxylamine (1). To a solution of
30 mg (0.083 mmol) of phenyl 9-triptycyl nitroxide (7) in 5 mL of
dichloromethane was added 9 mL (0.09 mmol) of phenylhydra-
zine, and the solution was stirred at room temperature for 5
min. Column chromatography on silica gel with dichlorometh-
ane–hexane (1:3) afforded 25 mg (0.069 mmol, 83%) of 1 as white
solids, mp 223–224 ꢁC (dec) (from dichloromethane–hexane).
Found: C, 86.61; H, 5.45; N, 3.85%. Calcd for C26H19NO: C,
1
86.40; H, 5.30; N, 3.88%. H NMR (CDCl3) ꢀ 5.381 (1H, s, 10-
H), 6.692 (1H, s, OH), 6.929 (3H, td, J ¼ 7:5, 1.4 Hz, 2/7/14-
H), 7.001 (3H, td, J ¼ 7:4, 1.3 Hz, 3/6/15-H), 7.089 (1H, m, p-
H), 7.20–7.26 (4H, m, o- and m-H), 7.413 (3H, dd, J ¼ 7:4, 1.4
Hz, 4/5/16-H), 7.451 (3H, d, J ¼ 7:5 Hz, 1/8/13-Hz). 13C NMR
(CDCl3) ꢀ 54.31 (1C, t), 76.17 (1C, q), 123.34 (3C, t), 123.76 (4C,
t), 124.31 (3C, t), 125.04 (3C, t), 125.15 (3C, t), 127.52 (1C, t),
143.05 (3C, q), 145.33 (3C, q), 149.13 (1C, q).
data with jIj > 2:0ꢁðIÞ were used. The function minimized was
2
ꢂwðjFoj ꢂ jFcjÞ , where w ¼ ½ꢁ2ðFoÞꢇꢂ1
.
Crystallographic data of these compounds have been deposited
at the CCDC, 12 Union Road, Cambridge, CB2 1EZ, UK and cop-
ies can be obtained on request, free of charge, by quoting the pub-
lication citation and the deposition numbers 254810–254814.
O-Ethyl-N-phenyl-N-9-triptycylhydroxylamine (2). To an
ice-chilled solution of 45 mg (0.124 mmol) of 1 in 8 mL of dry
Table 4. Crystal Data of Compounds 2, 4, 6, 7, and 8 and Parameters for Data Collection, Structure Determination,
and Refinement
Compound
2
4
6
7
8
Empirical formula
Formula weight
Crystal system
Space group
C28H23NO
389.47
monoclinic
P21=c
14.399(4)
16.953(5)
16.788(5)
91.7042(13)
4096.2(19)
8
1.263
1648
0.71070
113(2)
55
C26H19N
345.45
orthorhombic
P212121
13.446(5)
15.611(3)
8.763(2)
90.00
C27H19N3O3
433.47
monoclinic
P21=a
14.129(5)
8.025(3)
18.620(7)
94.017(7)
2106.1(13)
4
1.265
796
0.71070
113(2)
55
C26H18NO
360.44
monoclinic
P21=n
9.717(3)
15.718(4)
12.023(3)
93.380(13)
1833.1(8)
4
1.306
756
0.71070
110(2)
55
C26H17NO
359.43
monoclinic
P21=a
13.694(5)
8.122(3)
16.889(6)
106.687(15)
1799.3(11)
4
1.327
752
0.71069
110(2)
55
ꢀ
a/A
ꢀ
b/A
ꢀ
c/A
ꢂ/ꢁ
ꢀ 3
V/A
Z
1839.4(9)
4
1.247
Dc/g cmꢂ3
Fð000Þ
728
ꢃ(Mo Kꢄ)/cmꢂ1
Temp/ꢁK
0.71069
293(2)
54.5
2ꢅmax
No. of reflections measured
/
Total
Unique
7867
6900
2409
1637
3095
2890
3996
3566
3959
3440
No. of refinement variables
Final R; Rw
GOF
544
0.0525; 0.1384
1.060
321
0.038; 0.082
1.072
304
0.063; 0.167
1.048
326
0.046; 0.121
1.095
322
0.063; 0.156
1.202
R ¼ ꢂjjFoj ꢂ jFcjj=ꢂjFoj, Rw on F2.

C. Agawa et al.
Bull. Chem. Soc. Jpn., 77, No. 12 (2004) 2281
Selected Lengths and Angles of Compound 6. Selected
and J. Ojima, Tetrahedron, 52, 12409 (1996).
G. Yamamoto, H. Higuchi, M. Yonebayashi, Y. Nabeta,
and J. Ojima, Chem. Lett., 1995, 853.
G. Yamamoto, K. Inoue, H. Higuchi, M. Yonebayashi, Y.
Nabeta, and J. Ojima, Bull. Chem. Soc. Jpn., 71, 1241 (1998).
G. Yamamoto, H. Murakami, N. Tsubai, and M. Mazaki,
Chem. Lett., 1997, 605.
6 G. Yamamoto, N. Tsubai, H. Murakami, and M. Mazaki,
Chem. Lett., 1997, 1295.
7 G. Yamamoto, F. Nakajo, N. Tsubai, H. Murakami, and M.
Mazaki, Bull. Chem. Soc. Jpn., 72, 2315 (1999).
8 G. Yamamoto, F. Nakajo, and M. Mazaki, Bull. Chem. Soc.
Jpn., 74, 1973 (2001).
9 G. Yamamoto, K. Kuwahara, and K. Inoue, Chem. Lett.,
1995, 351.
10 G. Yamamoto, C. Agawa, and M. Minoura, Bull. Chem.
Soc. Jpn., 76, 825 (2003).
11 G. Yamamoto, F. Nakajo, N. Endo, and M. Mazaki, Chem.
Lett., 2000, 1102.
12 G. Yamamoto, F. Nakajo, N. Endo, and M. Mazaki, Bull.
Chem. Soc. Jpn., 74, 1467 (2001).
13 F. M. Logullo, A. H. Seitz, and L. Friedman, Org. Synth.,
Coll. Vol. V, 54 (1973).
ꢁ
ꢀ
lengths (A) and angles ( ) are as follows (the numberings are
shown in Fig. 1). N1–O1: 1.293(3), N1–N2: 1.278(3), N2–N3:
1.333(3), N1–C1: 1.478(3), C1–C2: 1.557(3), C1–C3: 1.547(3),
C1–C4: 1.536(4), N3–C5: 1.393(3), C6–O2: 1.211(4), C6–O3:
1.330(4), O1–N1–C1: 118.0(2), O1–N1–N2: 122.3(2), C1–N1–
N2: 118.7(2), N1–C1–C2: 117.8(2), N1–C1–C3: 108.6(2), N1–
C1–C4: 111.6(2), N1–N2–N3: 112.0(2), N2–N3–C5: 118.7(2),
N3–C5–C6: 120.0(2), C5–C6–C7: 120.3(2), C6–C7–O2:
124.9(3), C6–C7–O3: 113.2(3), C1–N1–N2–N3: 172.5(2), O1–
N1–N2–N3: 4.5(4), N1–N2–N3–C5: 171.2(2), N2–N1–C1–C2:
20.0(3), N2–N1–C1–C3: ꢂ101:2ð3Þ, N2–N1–C1–C4: 140.9(2),
O1–N1–C1–C2: ꢂ171:4ð2Þ, O1–N1–C1–C3: 67.4(3), O1–N1–
C1–C4: ꢂ50:5ð3Þ, N2–N3–C5–C6: ꢂ168:5ð3Þ, N3–C5–C6–C7:
ꢂ0:2ð4Þ, C5–C6–C7–O2: ꢂ14:7ð5Þ, C5–C6–C7–O3: 164.2(3).
Lineshape Analysis. Total lineshape analysis was performed
by visual matching of experimental spectra with theoretical spec-
tra computed on an NEC PC9821Xs personal computer equipped
with a Mutoh PP-210 plotter. For the calculations of theoretical
spectra, we used the DNMR3K program, a modified version of
the DNMR3 program27 converted for use on personal computers
by Dr. Hiroshi Kihara. Temperature dependences of chemical shift
differences and T2 values were properly taken into account.
Saturation Transfer Experiments on 8. In the study of the
Tp–N rotation, the 1,8-H signal at ꢀ 6.88 was saturated by irradi-
ation of a high-frequency field and the 13-H signal at ꢀ 8.12 was
monitored. The change in the signal intensity, IðtÞ, was followed
as a function of the duration of the irradiation, t, over the range
of 1 ms to 8 s. The data were fitted to the exponential function
(Eq. 1) and the three coefficients, A, B, and C, were determined
by nonlinear least squares analysis.28
3
4
5
14 T. C. Bissot, R. W. Parry, and D. H. Campbell, J. Am.
Chem. Soc., 79, 796 (1957).
15 D. H. Everett and W. F. K. Wynne-Jones, Proc. R. Soc.
London, A177, 499 (1941).
16 A. K. Majumdar and S. C. Saha, Anal. Chim. Acta, 40, 299
(1968).
17 S. B. Sarkar, M. Khalil, S. C. Saha, and S. K. Talapatra,
Acta Crystallogr., C39, 1075 (1983).
18 H. Kaur, M. J. Perkins, A. Scheffer, and D. C. Vennor-
Morris, Can. J. Chem., 60, 1594 (1982).
19 J. Ohshita, T. Iida, N. Ohta, K. Komaguchi, M. Shiotani,
and A. Kunai, Org. Lett., 4, 403 (2002).
IðtÞ
¼ A þ B expðꢂCtÞ
ð1Þ
Ið0Þ
ꢆ1
T1
ꢆ1
ꢆ
1
1
1
where A ¼
;
B ¼
;
C ¼
¼
þ
:
ꢆ1
ꢆ
T1
20 N. Benfaremo, M. Steenbock, M. Klapper, K. Mullen, V.
¨
Enkelmann, and K. Cabrera, Liebigs Ann. Chem., 1996, 1413.
21 W. B. Jennings and V. E. Wilson, ‘‘Acyclic Organonitro-
gen Stereodynamics,’’ ed by J. B. Lambert and Y. Takeuchi,
VCH, New York (1992), Chap. 6; H. O. Kalinowski and H.
Kessler, Top. Stereochem., 7, 295 (1973); A. P. Avedeenko, Russ.
J. Org. Chem., 36, 522 (2000).
Here T1 and ꢆ are the spin–lattice relaxation time and the
lifetime, respectively, of the 13-H signal. The rate constant for
the Tp–N rotation is given by 1=ð2ꢆÞ, because 1=ꢆ corresponds
to the sum of two rate constants, kð13-H ! 1-HÞ and kð13-H !
8-HÞ, and the rate constant for the Tp–N rotation corresponds to
one of them.
22 W. B. Jennings and D. R. Boyd, J. Am. Chem. Soc., 94,
7187 (1972).
23 A. Rieker and H. Kessler, Tetrahedron, 23, 3723 (1967).
Similarly, the rate constants for the C=N bond topomerization
were obtained by following the change in the signal intensity of
HD (ꢀ 7.81) upon irradiation of HB (ꢀ 6.57). Here the topomeriza-
tion rate constant is given by 1=ꢆ.
¯
24 G. Yamamoto and M. Oki, Bull. Chem. Soc. Jpn., 57, 2219
(1984).
25 W. Theilacker and K.-H. Beyer, Chem. Ber., 94, 2968
(1961).
26 G. M. Sheldrick, ‘‘Program for the Refinement of Crystal
We thank Professor Hideyo Matsuzawa of Kitasato Univer-
sity for obtaining the ESR spectrum of compound 7.
Structures,’’ University of Gottingen, Germany (1997).
¨
27 D. A. Kleier and G. Binsch, QCPE Program No. 165.
References
´
28 S. Forsen and R. A. Hofmann, J. Chem. Phys., 39, 2892
(1963); J. Sandstrom, ‘‘Dynamic NMR Spectroscopy,’’ Academic
¨
Press, London (1982), Chap. 4.
1
G. Yamamoto, H. Higuchi, M. Yonebayashi, and J. Ojima,
Chem. Lett., 1994, 1911.
G. Yamamoto, H. Higuchi, M. Yonebayashi, Y. Nabeta,
2