Unambiguous detection of 2,4,6-cycloheptatrien-1-ol by NMR spectroscopy and trapping with phenyltriazolinedione

Article abstract of DOI:10.1016/S0040-4020(02)01617-4

For the first time, the title alcohol 3 was unambiguously detected as a transient intermediate leading to ditropylether (6) in a cation-anion reaction between the tropylium ion (5) and hydroxide ion in an aqueous solution by NMR spectroscopy. It was also found that 6 gives 3 as a transient product in acid-catalyzed disproportionation of 6 into a mixture of tropone (4) and 1,3,5-cycloheptatriene (7) in chloroform containing water. Furthermore the existence of 3 in the latter reaction was confirmed by trapping with 4-phenyl-1,2,4-triazoline-3,5-dione to afford the [4+2]cycloadduct of the norcaradiene form of 3.

Full text of DOI:10.1016/S0040-4020(02)01617-4

Tetrahedron 59 (2003) 795–800
Unambiguous detection of 2,4,6-cycloheptatrien-1-ol by NMR
spectroscopy and trapping with phenyltriazolinedione
*
Mitsunori Oda, Kazuya Okawa, Hitoshi Tsuri and Shigeyasu Kuroda
Department of Applied Chemistry, Faculty of Engineering, Toyama University, Gofuku 3190, Toyama 930-8555, Japan
Received 30 October 2002; accepted 9 December 2002
Abstract—For the first time, the title alcohol 3 was unambiguously detected as a transient intermediate leading to ditropylether (6) in a
cation–anion reaction between the tropylium ion (5) and hydroxide ion in an aqueous solution by NMR spectroscopy. It was also found that 6
gives 3 as a transient product in acid-catalyzed disproportionation of 6 into a mixture of tropone (4) and 1,3,5-cycloheptatriene (7) in
chloroform containing water. Furthermore the existence of 3 in the latter reaction was confirmed by trapping with 4-phenyl-1,2,4-triazoline-
3,5-dione to afford the [4þ2]cycloadduct of the norcaradiene form of 3. q 2003 Elsevier Science Ltd. All rights reserved.
The thermodynamic stability of carbocations can be
evaluated by their pKRþ values. One method for determin-
ing the value is based on pH dependence of the electronic
absorption spectrum, assuming the equilibrium between a
carbocation and its corresponding alcohol (Eq. (1)).1 While
the alcohol corresponding to a thermodynamically unstable
carbocation is stable enough to be isolated, the alcohol
corresponding to a thermodynamically stable cation is
basically unstable because of its facile dissociation back to
the cation. In some cases, its easy and irreversible
disproportionation to the correspnding ketone and hydro-
carbon species is observed. For example, 1-phenalenol (1)
had been elusive until Sugihara and Murata found special
protocols for isolating it in 1989;2 they used sodium
borohydride in the presence of ceric chloride for reduction
of phenalenone (2) and isolated 1 by particular workup and
crystallization. However, they reported that application of
their method was unsuccessful in isolation of the title
alcohol, 2,4,6-cycloheptatrien-1-ol (3),3,4 from tropone (4).5
Although the equilibrium between the tropylium ion (5) and
the alcohol 3 (in Eq. (1); R¼2,4,6-cycloheptatrien-1-yl) was
postulated, particularly well in pKRþ value measurements of
5, such as in the pioneering work by Doering et al. in 1954,6
3 has, surprisingly, been neither observed nor isolated yet in
the half century since the definite finding of the tropylium
ion. Doering et al. found that 5 affords only ditropylether (6)
as a single isolable product under the basic aqueous
conditions. Later on, it was also found that disproportiona-
tion of 6 under acidic conditions gave tropone (4) and 1,3,5-
cycloheptatriene (7),7 that makes it harder to detect 3
(Schemes 1–3).
ð1Þ
Meanwhile, in 1976 Franzus et al. reported that a base-
catalyzed rearrangement of 7-norbonadienol (8) provides
the tropyl skeleton.8 They suggested that the first product
observed in the rearrangement under basic conditions could
be the alkoxide of 3 by NMR analysis. However, signals in
the 60 MHz NMR spectrum can be assigned either to those
of the alkoxide or to those of 6, as pointed out by the authors
themselves. It is notorious that the signals did not show
time-dependent change after the disappearence of 8. Thus, it
is hard to concede that definite detection of either 3
equilibrated with its alkoxide or the alkoxide itself is
achieved in the study of Franzus et al. In this paper we
describe the unambiguous detection of 3 as a transient
species under acidic and basic conditions by NMR
spectroscopy and its trapping with phenyltriazolinedione
as a dienophile (Scheme 4).
Scheme 1.
1. Results and discussion
1.1. Detection of 3 in the reaction of the tropylium cation
and hydroxide anion
Keywords: carbocation; alcohol; tropylium ion; cycloaddition;
phenyltriazolinedione.
*
Corresponding author. Tel.: þ81-76-445-6820; fax: þ81-76-445-6819;
Our approach to detection of the title alcohol 3 was
e-mail: oda@eng.toyama-u.ac.jp
0040–4020/03/$ - see front matter q 2003 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(02)01617-4

796
M. Oda et al. / Tetrahedron 59 (2003) 795–800
Scheme 2.
of about 40 min (Fig. 1). After then the spectrum showed
only the signals of 6 and the resulted mixture looked cloudy
because of its low solubility in the aqueous solution. The
latter transient signals disappeared with a half life-time of
4 min under the conditions and with the slightly longer half
life-time of 7 min at 78C. The 1H NMR spectrum after 2 min
is shown in Figure 2. The signals consist of four different
protons with their integrals of 1:2:2:2 ratio, the same as seen
in that of 6. This splitting pattern is similar to that of 6, but
the chemical shifts are different from those of 6, clearly
indicating the existence of a new 7-substituted 1,3,5-
cycloheptatriene. Based on the integral pattern and the
chemical shifts observed, this fleeting intermediate is best
regarded as the title alcohol 3. The signals of 3 appear
broadened probably due to slow equilibrium between 3 and
5, because unimolecular dynamic processes for 3, such as
ring inversion and norcaradiene-cycloheptariene tautomer-
ization, and equilibrium with the corresponding alkoxide are
expected to be very fast at the NMR temperature.
Monitoring the reaction by 13C NMR spectroscopy also
supports the existence of an intermediate into 6 and its
Scheme 3.
straightforward. Initially we monitored the reaction of the
tropylium cation and sodium hydroxide in deterium oxide
by NMR spectroscopy. Fortunately, we were able to achieve
our purpose by this simple experiment which has not
appeared in the literature. Tropylium tetrafluoroborate was
added to a solution of 2 equiv. of NaOD in D2O in an NMR
1
tube and the reaction mixture was monitored by H NMR
analysis with sodium 4,4-dimethyl-4-silapentanesulfonate
(DSS) as an internal standard at 238C of the probe
temperature. At the beginning of the NMR measurements,
the signals of 5 had already disappeared. Growing signals
for 6 and decreasing signals were observed in a time range
O—
O—
O—
6
8
Scheme 4.
Figure 1. Change in time-dependent product ratio in the reaction of the tropylium ion (5) and sodium deuteroxide in D2O at 238C. The ordinate shows the molar
ratios of 2,4,6-cycloheptatrien-1-ol (3, W) and ditropylether (6, A) to the internal standard (DSS) and the abscissa time in min.

M. Oda et al. / Tetrahedron 59 (2003) 795–800
797
Figure 2. The 1H NMR spectrum of the reaction mixture of the tropylium ion (5) and sodium deuteoxide in deuterium hydroxide after 2 min at 238C.
structure as 3. The assignment of carbon and proton signals
for 3, 6, and 7-methoxy-1,3,5-cycloheptatriene (9) for
comparison are listed in Figure 3. Chemical shifts of the
hydrogens at the 1–3 positions of 3 were observed at higher
magnetic field. Particularly, the hyrogen of the 1 position of
3 resonated relatively at higher field by more than 0.5 ppm
compared with those at the corresponding position of 6 and
9. This may be attributed to the effect of the electronic
nature of the alkoxide which naturally exists in equilibrium
with 3 under basic conditions. The final product 6 was
isolated from this reaction mixture in more than 90% yield
as reported.5
1.2. Detection of 3 in acid-catalyzed disproportionation
reaction of 6 and trapping with 4-phenyl-1,2,4-tri-
azoline-3,5-dione as a dienophile
Encouraged by the experiments for the detection of 3 from
5, we proceeded to detect 3 in acid-catalyzed disproportio-
nation7 of ditropylether (6) into a mixture of tropone (4) and
1,3,5-cycloheptatriene (7). A reaction mixture of 6 and
trifluoroacetic acid (TFA, 6 equiv. to 6) in CDCl3 saturated
with D2O was monitored at 238C. As the signals of 6
decreased, the signals of both 4 and 7 increased in the course
of reaction time. Along with these signal changes, it was
observed that another set of signals grew up to half of the
reaction period and then gradually disappeared (Fig. 4).
After 65 min, the reaction provided a quatitative yield of a
1:1 mixture of 4 and 7. The transient product observed in the
Figure 3. Proton and carbon chemical shifts of 2,4,6-cycloheptatrien-1-ol
(3) and ditropylether (6) in NaOD-D2O and 7-methoxy-1,3,5-cyclo-
heptatriene (9) in D2O with DSS as an internal standard.
1
reaction can be regarded best also as 3 from its H NMR

798
M. Oda et al. / Tetrahedron 59 (2003) 795–800
Figure 4. Change in time-dependent product ratios in the disproportionation of ditropylether (6) in CDCl3–CF3CO2D at 238C. The ordinate shows the molar
ratios of 2,4,6-cycloheptatrien-1-ol (3, W), ditropylether (6, A), tropone (4, X), and 1,3,5-cycloheptatriene (7, K) to the internal standard (TMS) and the
abscissa time in min.
signals in the mixture. While all proton signals of 6 in acidic
CDCl3 were observed at slightly higher field than those in
basic D2O, all proton signals of 3 in acidic CDCl3 were
observed at lower field than those in basic D2O (Figs. 3 and
5). In this case also, the hydrogen at the 1 position shifts
most to lower field, probably due to more effective
protonation of the hydroxy group of 3 than the ether
functional group. The proton chemical shifts of 6 and 9
seem to be affected mainly by solvent polarity. Thus, the
greater dielectric constant of the solvent used, the lower
chemical shifts of hydrogens of 6 and 9 were observed. On
the other hand, the proton chemical shifts of 6 and 9 seem to
be affected rather by pH of the solution.
conditions with undeuterated solvents, we succeeded in
trapping with 4-phenyl-1,2,4-triazoline-3,5-dione
3
(PTAD). The PTAD adduct 10, to the norcaradiene form
of 3, was obtained in 72% yield based on twice amounts of 6
used, accompanied with a mixture of previously known
adducts 119 and 1210 in 22 and 21% yields based on the
amount of 6 used, respectively. The stereochemistry of the
hydroxy group on the cyclopropane ring in 10 was assigned
to be exo based on the vicinal coupling constant
(3JH–H¼1.2 Hz) between the hydrogen on the carbon
atom attached by the hydroxy group and the neighboring
methine hydrogens on the cyclopropane ring.11 The yield of
10 decreased in higher concentration of TFA and increased
in its lower concentration as shown in Table 1. Nozoe has
previously proposed two possible pathways from 6 into a
mixture of 4 and 7;7acid-catalyzed cleavage of the ether
linkage of 6 provides 3 and 5 and subsequent intermolecular
hydride shift from 3 into either 5 or 6 ends up with a mixture
of 4 and 7. The dependence of the yields on the amount of
the acid in the trapping experiments clearly indicates that
the acid favors consumption of 3 rather than its accumu-
lation, assuming that the adduct formation step is not
affected by the acid. Although the definite mechanism of the
formation of 4 and 7 must await further meticulous kinetic
study, the existence of 3 under these reaction conditions is
evident (Scheme 5).
Further confirmation of the existence of 3 was made by its
trapping with a dienophile. Under the same reaction
Table 1. Result of trapping experiments of 3 with PTAD under the
conditions with various concentrations of TFA
TFA (molar equiv. to 3)
Yield of the adducts (%)
10a
11b
12b
20
5
1
16
65
72
81
27
22
77
25
21
a
Molar ratio based on twice amounts of 3 used.
Molar ratio based on the amount of 3 used.
b
2. Summary
For the first time, we have observed the title alcohol, 2,4,6-
cycloheptatrien-1-ol (3), as a transient species by NMR
spectroscopy in the cation–anion reaction. The half-life
Figure 5. Proton chemical shifts (d ppm) of tropyl derivatives in CDCl3.
The values in parentheses were observed in the disproportionation mixture
in CDCl3–CF3CO2D. TMS was used as an internal standard.

M. Oda et al. / Tetrahedron 59 (2003) 795–800
799
Scheme 5.
1
time of 3 in this reaction was found to be in order of several
minutes. Also, 3 was detected in disproportionation of
ditropylether (6). In the latter reaction in the presence of
PTAD, 3 was trapped as a [4þ2]cycloadduct of its
norcaradiene form to give the adduct 10 along with
previously known 11 and 12. The stereochemistry of the
hydroxy group at the cyclopropane ring in 10 was found to
be exo. The yield of 10 in the trapping experiments was
found to be dependent on the concentration of TFA. These
results will help in elucidating the definite mechanism of the
disproportionation of 6 which is a basic reaction in
troponoid chemistry.
controlled temperature of 78C by H NMR spectroscopy.
The same procedure was repeated for monitoring at 238C by
13C NMR spectroscopy. After the monitoring, the resulted
mixture of one experiment was poured into water and
extracted with hexane (5£10 ml). The combined organic
layer was dried with MgSO4 and the solvent was removed to
give 2.6 mg (94%) of 66,7 as a brown oil.
3.3. Detection of 2,4,6-cycloheptatrien-1-ol (3) in the
disproportionation of the ditropylether (6)
In an NMR tube was charged a solution of 10.0 mg
(50.1 mmol) of 6 and 5mg of DSS in 0.60 ml of chloroform-
d. To this solution was added 50 ml (300 mmol) of TFA. The
reaction mixture was monitored by H NMR spectroscopy
at 238C.
1
3. Experimental
3.1. General
After 3 h, the resulted mixture was poured into a phosphate
buffer (pH 6.86) and extracted with dichloromethane
(5£10 ml). The combined organic layer was dried with
MgSO4 and the solvent was removed to give 9.6 mg (96%)
of a 1:1 mixture of 4 and 7 as a yellow oil.
Melting points were measured on a Yanaco MP-3 and are
uncorrected. IR spectra were recorded on a JASCO IR-810
and Perkin–Elmer Spectrum RX I spectrometers. UV
spectra were measured on a Shimadzu UV-1600 spec-
trometer. 1H NMR (400 MHz) and 13C NMR (100 Hz) were
recorded on a JEOL a400. Mass spectra were measured on a
JEOL GC-Mate mass spectrometer. Column chromato-
graphy was done with Merck Kieselgel 60 Art 7734.
Deuterium oxide (isotope purity; .99.9%) from Sigma-
Aldrich Co., chloroform-d (isotope purity; .99.5%) from
Wako Co. and 40% (w/w) sodium deuteroxide in deuterium
oxide (isotope purity; 99%) from Merck Inc. were used
without purification. DSS was purchased from Merck Inc.
Tropylium tetrafluoroborate was prepared by the method of
Dauben et al.12 and was purified by recrystallization from a
mixture of dichloromethane and hexane. PTAD was
prepared according to the method in the literature.13
3.4. Trapping experiments of 2,4,6-cycloheptatrien-1-ol
(3) with PTAD in the disproportionation of the
ditropylether (6)
A solution of 6 (45.5 mg, 0.230 mmol) in in 2 ml of
chloroform was added 5 ml of water and 18 ml (0.23 mmol)
of TFA, followed by addition of 95.2 mg (0.460 mmol) of
PTAD. After being stirred for 5 h, the reaction mixture was
concentrated in vacuo. The residue was purified by silica gel
chromatography with a 1:1 mixture of chloroform and ethyl
acetate as eluent to give 12, 11, and 10 in this order. The
yields in the reactions with other TFA concentrarions are
shown in Table 1.
3.2. Detection of 2,4,6-cycloheptatrien-1-ol (3) in the
reaction of the tropylium cation and hydroxide anion
3.4.1. Compound 12. 12.7 mg, 21%; colorless micro-
crystals, mp¼199–2008C (lit.10 186–1888C). 1H NMR
(CDCl3) d¼0.32 (dt, J¼6.6, 3.8 Hz, 1H), 0.67 (q, J¼7.1 Hz,
1H), 1.61 (m, 2H), 5.21 (m, 2H), 6.10 (m, 2H), 7.32–7.45
(m, 5H); 13C NMR (CDCl3) d¼6.22, 8.40, 53.66, 125.18,
125.56, 128.24, 129.11, 131.42, 156.71.
In an NMR tube was charged 0.60 ml of a 0.1 M deuterium
hydroxide solution and 2 mg of DSS in deuterium oxide. To
this solution was added 5.0 mg (28 mmol) of tropylium
tetrafluoroborate in one portion. The NMR tube was quickly
shaken and was loaded into the NMR probe. The reaction
was monitored at the probe temperature of 238C or the
3.4.2. Compound 11. 14.2 mg, 22%; yellow prisms,

800
M. Oda et al. / Tetrahedron 59 (2003) 795–800
mp¼169–1708C (lit.9 178–1808C). 1H NMR (CDCl3)
d¼5.38 (dd, J¼7.3, 2.0 Hz, 1H), 5.44 (t-like, J¼7.3 Hz,
1H), 6.02 (dd, J¼10.9, 2.0 Hz, 1H), 6.93 (t-like, J¼8.1 Hz,
1H), 7.11 (dd, J¼7.8, 7.6 Hz, 1H), 7.35–7.51 (m, 5H); 13C
NMR (CDCl3) d¼51.38, 63.55, 124.28, 125.46, 128.52,
129.22, 130.88, 131.17, 136.23, 143.35, 151.77, 152.39,
188.47.
Eds.; Interscience: New York, 1968; Vol. 1, pp 77–110,
Chapter 2.
2. Sugihara, Y.; Kawanaka, H.; Murata, I. Angew. Chem. 1989,
101, 1258–1259, Angew. Chem. Int. Ed. 1989, 28, 1258–
1269.
3. For isomeric enol-type compounds, 1,3,5-cycloheptatrien-1-,
-2-, and -3-ols, see: Lew, C. S. Q.; Capon, B. J. Org. Chem.
1997, 62, 5344–5353.
3.4.3. Compound 10. 87.6 mg, 72%; colorless micro-
1
4. Although an iron tricaronyl complex of 3 has been known,
cleavage of iron from this complex for either detection or
trapping was so far unsuccessfull despite our efforts. For the
synthesis of this complex, see: Mayr, H.; Mu¨ller, K. H.; Ofinal,
A. R.; Bu¨hl, M. J. Am. Chem. Soc. 1999, 121, 2418–2424.
5. Sugihara, Y.; Murata, I. Private communications.
6. Doering, W. v. E.; Knox, L. H. J. Am. Chem. Soc. 1954, 76,
3203–3206. Doering, W. v. E.; Knox, L. H. J. Am. Chem. Soc.
1957, 79, 352–356.
crystals, mp¼202–2038C. H NMR (CDCl3) d¼1.54 (s,
1H), 1.88 (m, 2H), 2.88 (t, J¼1.2 Hz, 1H), 5.24 (m, 2 H),
6.21 (t-like, J¼3.7 Hz, 2H), 7.33–7.47 (m, 5H); 13C NMR
(CDCl3) d¼13.37, 51.41, 60.17, 125.51, 126.49, 128.40,
129.17, 131.19, 156.59. IR (KBr) n¼3449brm, 1770m,
1713s, 1500m, 1457w, 1403s, 1233m, 1160m, 1138m,
1045m, 1022m, 899w, 867w, 843w, 773m, 736m, 720m,
640w cm21; MS (70 eV) m/z (rel intensity) 283 (Mþ, 0.5),
177 (23), 119 (41), 106 (30), 105 (30), 91 (100), 77 (50); UV
7. Ikemi, T.; Nozoe, T.; Sugiyama, H. Chem. Ind. (London) 1960,
923. Borg, A. P. T.; Helden, R. V.; Bickel, A. F.; Renold, W.;
Dreiding, A. S. Helv. Chim. Acta 1960, 457–458. Machiguchi,
T. Synth. Commun. 12, 1982, 1021–1025.
l
max (CH2Cl2) 230 nm (log 1¼3.99), 256 (3.60). Found: C,
63.78; H, 4.70; N, 14.93%. Calcd for C15H13N3O3: C, 63.60;
H, 4.63; N, 14.83%.
8. Franzus, B.; Shceinbaum, M. L.; Waters, D. L.; Bowlin, H. B.
J. Am. Chem. Soc. 1976, 98, 1241–1247.
Acknowledgements
9. Sasaki, T.; Kanematsu, K.; Hayakawa, K. J. Chem. Soc. (C)
1971, 2142–2147.
The present work is supported by the Grant-in-Aid for
Scientific Research (No. 13640528 to M. O.) from the
Ministry of Education, Culture, Sports, Science, and
Technology, Japan. We thank Mr T. Fujimura at Toyama
University for the technical assistance at the beginning of
this study and also Dr R. Miyatake at Toyama University for
help in preparing manuscript of this paper. We also thank
Professor Y. Sugihara at Yamaguchi University and
Professor T. Machiguchi at Saitama University for helpful
discussions.
10. Cookson, R. C.; Gilani, S. S. H.; Stevens, I. D. R. J. Chem.
Soc. (C) 1967, 1905–1909.
11. Adam, et al. prepared the exo-adduct of 7-methoxycyclo-
heptatriene having a similar small coupling constant of 1.4 Hz
Adam, W.; Balci, M.; Pietrzak, B. J. Am. Chem. Soc 1979,
101, 6285–6291.
12. Dauben, Jr. H. J.; Gadecki, P. A.; Harman, M. K.; Peason, D. L.
J. Am. Chem. Soc. 1957, 1957–1958.
13. Cookson, R. C.; Gupte, S. S.; Stevens, I. D. R.; Watts, C. T.
Organic Synthesis; Wiley: New York, 1988; Coll. Vol. 6,
pp 936–940.
References
1. Franklin, J. L. Carbonium Ions; Olah, G. H., Schleyer, P. v. R.,