Solvolytic generation of antiaromatic cyclopentadienyl cations

Article abstract of DOI:10.1021/ja962494z

Solvolysis of 1,3-di-tert-butyl-5-methyl-5-cyclopenta-1,3-dienyl trifluoroacetate (15a) occurs with a strong dependence on solvent ionizing power (m = 0.97), and gives products of substitution, allylic and skeletal rearrangement with substitution, and eli

Full text of DOI:10.1021/ja962494z

J. Am. Chem. Soc. 1997, 119, 2371-2375
2371
Solvolytic Generation of Antiaromatic Cyclopentadienyl Cations
Annete D. Allen, Mirjana Sumonja, and Thomas T. Tidwell*
Contribution from the Department of Chemistry, UniVersity of Toronto,
Toronto, Ontario, Canada M5S 3H6
ReceiVed July 18, 1996. ReVised Manuscript ReceiVed December 2, 1996X
Abstract: Solvolysis of 1,3-di-tert-butyl-5-methyl-5-cyclopenta-1,3-dienyl trifluoroacetate (15a) occurs with a strong
dependence on solvent ionizing power (m ) 0.97), and gives products of substitution, allylic and skeletal rearrangement
with substitution, and elimination. These results are characteristic of a process involving an intermediate carbocation,
and provide the first measurements of the kinetics of formation of a cyclopentadienyl cation, the prototypical 4π-
electron carbocation destabilized by antiaromaticity. The reactivity of 15a in 2,2,2-trifluoroethanol at 25 °C is
calculated to be lower than those of analogous fluorenyl and indenyl derivatives by factors of 3 × 104 and 4 × 102,
and is exceeded by that calculated for 1,3-dimethyl-3-cyclopentenyl trifluoroacetate by a factor of 1014, showing the
large carbocation destabilizing effects of antiaromaticity.
The cyclopentadienyl cation (1) is the prototypical carbocation
destabilized by 4π-electron antiaromaticity,1 and has long been
the subject of study.2,3 These investigations include the
4 of 56.7 kcal/mol for the reaction of eq 2 both indicate major
antiaromaticity effects.
+
electrochemical determination of the pKR as -40 or lower,
which is 20 units lower than those for representative conjugated
cations,2c observation of C5H5+ as a triplet with the D5h structure
1 in a matrix of SbF5 at 78 K,2d and the failure of cyclopenta-
+
dienyl iodide to form C5H5 solvolytically even when treated
with silver perchlorate in propionic acid at -15 °C (eq 1).2b
By contrast 2 is 10 times more reactive than cyclopentyl iodide
in bimolecular reaction with bromide.2e
These studies indicate that cyclopentadienyl carbocations are
highly destabilized and that their generation under solvolytic
conditions might be considered unattainable, but there are some
recent indications that such species may indeed be generated
in solution. Thus, the reaction of pentamethylcyclopentadienyl
bromide (6) with silver tetrafluoroborate at -10 °C in CH2Cl2
with methanol, methylamine, dimethylamine, pyridine, and
dimethyl sulfide led to substitution products, which could result
by capture of the pentamethylcyclopentadienyl cation (7) by
the nucleophiles (eq 3).4a Also the reaction of 5-tolyl-1,2,3,4-
tetraphenylcyclopentadienyl bromide (8) with substituted silver
acetates gave a nonequilibrium mixture of acetate esters 10,
presumably through the cation 9 (eq 4).4b Upon heating in the
range of 50-150 °C the acetates 10 underwent equilibration in
a process which was catalyzed by Hg(II) and Pd(II) salts, and
this transformation was ascribed to 3,3-sigmatropic rearrange-
ments of the acyloxy groups.4b
Despite these extensive previous studies of cyclopentadienyl
cations, and speculations “that rate-accelerating substituent
effects may be used for increasing the reactivity of cyclopen-
tadienyl precursors into an experimentally convenient range”,4c
there have been no reported measurements of the kinetics of
the solvolytic reactivity of cyclopentadienyl derivatives leading
to cyclopentadienyl cations. We have been engaged in the study
+
Carbon scrambling of C5H5 in the gas phase is suggested
to involve pyramidal ions 3,2f although the singlet ion is
calculated3c to prefer C2V ethylene-type structures 4 which
interconvert with an extemely low 0.09 kcal/mol barrier through
C2V allylic-type structures 5. The calculated3d magnetic sus-
ceptibility exaltation and the homodesmotic destabilization of
* To whom correspondence should be addressed. E-mail: tidwell@
scar.utoronto.ca. Fax: (416) 287-7204. Phone: (416) 287-7217.
X Abstract published in AdVance ACS Abstracts, February 15, 1997.
(1) (a) Tidwell, T. T. Angew. Chem., Int. Ed. Engl. 1984, 23, 20-34.
(b) Minkin, V. I.; Glukhovtsev, M. N.; Simkin, B. Ya. Aromaticity and
Antiaromaticity; Wiley-Interscience: New York, 1994.
(2) (a) Breslow, R. Acc. Chem. Res. 1973, 6, 393-398. (b) Breslow, R.;
Hoffman, J. M., Jr. J. Am. Chem. Soc. 1972, 94, 2110-2111. (c) Breslow,
R.; Mazur, S. J. Am. Chem. Soc. 1973, 95, 584-585. (d) Saunders, M.;
Berger, R.; Jaffe, A.; McBride, J. M.; O’Neill, J.; Breslow, R.; Hoffman, J.
M., Jr.; Perchonock, C.; Wasserman, E.; Hutton, R. S.; Kuck, V. J. J. Am.
Chem. Soc. 1973, 95, 3017-3108. (e) Breslow, R.; Canary, J. W. J. Am.
Chem. Soc. 1991, 113, 3950-3952. (f) Schwarz, H.; Thies, H.; Franke, W.
In Ionic Processes in the Gas Phase; Ferreira, M. A. A., Ed.; Reidel Co.:
Dordrecht, The Netherlands, 1984.
(3) (a) Ko¨hler, H.-J.; Lischka, H. J. Am. Chem. Soc. 1979, 101, 3479-
3486. (b) Feng, J.; Leszczynski, J.; Weiner, B.; Zerner, M. C. J. Am. Chem.
Soc. 1989, 111, 4648-4655. (c) Glukhovtsev, M. N.; Reindl, B.; Schleyer,
P. v. R. MendeleeV Commun. 1993, 100-102. (d) Schleyer, P. v. R.;
Freeman, P. K.; Jiao, H.; Goldfuss, B. Angew. Chem., Int. Ed. Engl. 1995,
34, 337-339. (e) Glukhovtsev, M. N.; Bach, R. D.; Laiter, S. J. Phys. Chem.
1996, 100, 10952-10955.
(4) (a) Jutzi, P.; Mix, A. Chem. Ber. 1992, 125, 951-954. (b) Dushenko,
G. A.; Mikhailov, I. E.; Kamenetskaya, I. A.; Skachkov, R. V.; Zhunke,
A.; Myugge, K.; Minkin, V. I. Russ. J. Org. Chem. 1994, 30, 1559-1564.
(c) Friedrich, E. C.; Tam, T. M. J. Org. Chem. 1982, 47, 315-319.
(5) (a) Allen, A. D.; Colomvakos, J. D.; Tee, O. S.; Tidwell, T. T. J.
Org. Chem. 1994, 59, 7185-7187. (b) Allen, A. D.; Mohammed, N.;
Tidwell, T. T. J. Org. Chem. 1997, 62, 246-252. (c) Allen, A. D.; Fujio,
M.; Tee, O. S.; Tidwell, T. T.; Tsuji, Y.; Tsuno, Y.; Yatsugi, K. J. Am.
Chem. Soc. 1995, 117, 8974-8981. (d) Kirmse, W.; Wonner, A.; Allen,
A. D.; Tidwell, T. T. J. Am. Chem. Soc. 1992, 114, 8828-8835.
S0002-7863(96)02494-8 CCC: $14.00 © 1997 American Chemical Society

2372 J. Am. Chem. Soc., Vol. 119, No. 10, 1997
Allen et al.
Table 1. Solvolytic Rate Constantsa for 5-Methyl- and
5-Phenyl-1,3-di-tert-butyl-5-cyclopenta-1,3-dienyl Trifluoroacetate
(15a,b)
kobsd (s-1 × 104)
T (°C) 5-Me (15a) 5-Ph (15b) kMe/kPh
solvent
YOTFA
97 HFIP
97 HFIPc
97 TFEc
3.37 (3.61)b 25
15.4
14.4
36.5
8.68
2.72
0.343
1.76
3.67
4.2
25
2.25 (1.83)b 70.1
15.8
4.23
2.3
2.1
2.9
2.5
54.4
43.6
25d,f
0.928
0.138e
80 EtOHc
0.0
76.0
64.8
0.453
0.204
0.00543
9.16
54.8
25.0d,g
62.6
60 MeOHc 1.50
(1.52)b
53.9
3.82
1.25
0.124
1.81
44.2
25.0d,h
80 MeOHc 0.63 (0.47)b 62.7
53.9
44.3
0.625
0.249
0.023
of the solvolytic generation of CF3-substituted doubly-destabi-
lized fluorenyl (11)5a and (12)5b indenyl carbocations, and other
destabilized carbocations,5c,d and now report the first kinetic
studies of cyclopentadienyl cation formation.
25.0d,i
a Duplicate runs at each temperature, (3%. YOTs c Containing 2
.
b
equiv of 2,6-lutidine. d Extrapolated from data at higher temperatures.
e ∆Hq ) 20.9 kcal/mol, ∆Sq ) -10.8 cal mol-1 K-1, log k ) 1.28
YOTFA -7.74. f ∆Hq ) 20.5 kcal/mol, ∆Sq ) -10.3 cal mol-1 K-1, log
k ) (0.97 ( 0.09) YOTFA - (6.33 ( 0.17). g ∆Hq ) 22.5 kcal/mol, ∆Sq
) -11.8 cal mol-1 K-1
.
h ∆Hq ) 22.3 kcal/mol, ∆Sq ) -6.2 cal mol-1
K-1
.
i ∆Hq ) 22.1 kcal/mol, ∆Sq ) 10.1 cal mol-1 K-1
.
Scheme 1
Results
Photolysis of commercially available 3,5-di-tert-butyl-o-
benzoquinone gives efficient formation of 2,4-di-tert-butylcy-
clopentadienone (13),6 which on reaction with methyllithium
or phenyllithium forms the alcohols 14a and 14b, respectively,
which are converted by trifluoroacetic anhydride to the respec-
tive trifluoroacetates 15 (eq 5). Trifluoroacetate leaving groups
under the conditions of these experiments, and this is due to
the presence of the bulky tert-butyl groups.6c
The reactions of 15a and 15b in 97% 1,1,1,3,3,3-hexafluoro-
2-propanol (HFIP) and 97% 2,2,2-trifluoroethanol (TFE), and
of 15a in 60% MeOH, 80% MeOH, and 80% EtOH, were
monitored by UV spectroscopy and took place with first-order
kinetics, as reported in Table 1. The rates of these reactions
gave correlations with the solvent parameters for trifluoroacetate
have been frequently exploited in studies of solvolysis.7 They
are easy to prepare even for crowded substrates, and have a
convenient reactivity intermediate between those of tosylate and
p-nitrobenzoate leaving groups.7 The substrates 14 and 15
showed no evidence for dimerization by [4 + 2] cycloaddition
7c,d
leaving groups YOTFA by the relationships log k ) (0.97 (
0.09)YOTFA - (6.33 ( 0.17) for 15a and log k ) 1.28YOTFA
-
(6) (a) DeSelms, R. C.; Schleigh, W. R. Synthesis 1973, 614-615. (b)
Alkenings, B.; Betterman, H.; Dasting, I.; Schroers, H.-J. Spectrochim. Acta
1993, 49A, 315-320. (c) Garbisch, E. W., Jr.; Sprecher, R. F. J. Am. Chem.
Soc. 1969, 91, 6785-6799.
(7) (a) Noyce, D. S.; Virgilio, J. A. J. Org. Chem. 1972, 37, 2643-
2647. (b) Lambert, J. B.; Tang, G.-t.; Finzel, R. B.; Teramura, D. H. J.
Am. Chem. Soc. 1987, 109, 7838-7845. (c) Bentley, T. W.; Roberts, K. J.
Chem. Soc., Perkin Trans. 2 1989, 1055-1060. (d) Bentley, T. W.;
Llewellyn, G. Prog. Phys. Org. Chem. 1990, 17, 121-158. (e) Creary, X.;
Wang, Y.-X. J. Org. Chem. 1992, 57, 4761-4765. (f) Creary, X.; Jiang,
Z. J. Org. Chem. 1994, 59, 5106-5108. (g) Creary, X.; Wang, Y.-X.; Jiang,
Z. J. Am. Chem. Soc. 1995, 117, 3044-3053. (h) Takeuchi, K.; Kitagawa,
T.; Ohga, Y.; Yoshida, M; Akiyama, F.; Tsugeno, A. J. Org. Chem. 1992,
57, 280-291.
7.74 for 15b. The YOTFA values7c,d for these solvents are rather
similar to the corresponding YOTs values, as noted in Table 1.
For 15a the use of YOTs values gives the correlation log k )
(0.93 ( 0.04)YOTs - (6.20 ( 0.08).
Preparative scale solvolysis of 15a in CF3CH2OH and
chromatographic separation of the products led to the isolation
of 16a, 17, 18, and 19 (Scheme 1), whose structures were
established by their spectroscopic properties as described below.
On the basis of integration of the 1H NMR signals for the crude
reaction mixture the relative yields of 16-19 (R1 ) CF3CH2)

Generation of Antiaromatic Cyclopentadienyl Cations
J. Am. Chem. Soc., Vol. 119, No. 10, 1997 2373
were 66, 10, 14, and 10%, respectively. To estimate the
comparative yields for solvolysis in C2D5OD, the product of a
small-scale run was analyzed by H NMR as containing 16a,
Scheme 2
1
17, and 19 (R1 ) C2D5) in relative percentages of 42, 40, and
18. Structures were assigned from the 1H NMR signals
corresponding to products identified in the study in CF3CH2-
OH. The products in C2D5OD were also analyzed by GC/MS
and assigned by the EIMS fragmentation patterns, and the
possible presence of the isomeric structure 16b was also noticed.
Determinations of the relative yields (parentheses) of products
for R1 ) C2D5 from the GC peak areas were 16a (43), 16b (3),
17 (32), 18 (<1), and 19 (17%), with 5% starting material, and
these are in reasonable agreement with the independent mea-
1
surement by H NMR. Thus, although the products for the
at C-5. The absence of visible nonequivalence of the diaste-
reotopic geminal methyl groups or of the CH2 protons of the
CH2CF3 group is also consistent with this structure as opposed
to an isomeric 1-methyl-3-tert-butyl structure, in which there
would be a greater probability of observable nonequivalence
because of the proximity of the CH3 at C-1.
reaction in C2D5OD were not isolated and positively identified,
there are indications that the pattern of products in this medium
is similar to that in TFE.
The identification of the unrearranged substitution product
1
19-OCH2CF3 follows from the close resemblance of the H
and 13C NMR spectra to those of 14a and 15a, including the
1H NMR absorption of the carbinyl methyl at δ 1.55 and the
2,4-vinyl hydrogens at δ 6.05 and 5.35, respectively, as
compared to values of 1.53, 5.87, and 5.48, respectively, for
14a, and 1.74, 6.06, and 5.75 for 15a. This compound also
showed a distinct nonequivalence of the diastereotopic CH2
protons of the CH2CF3 group, with the absorption of the protons
as two overlapping quartets at δ 3.503 and 3.513 (JH,F ) 8.9
Hz). The structure of the product of allylic rearrangement 16a
Discussion
The kinetic data for both substrates 15a and 15b show a
strong dependence on solvent polarity, with 3000-fold greater
reactivity for 15a in 97% HFIP compared to 80% EtOH at 25
°C, and m values of 0.97 and 1.28, respectively. These large
values are diagnostic that the reactions occur by carbocationic
processes. This conclusion is consistent with the finding that
the reaction products from 15a in TFE and EtOH represent
elimination, substitution, substitution with allylic rearrangement,
and substitution with skeletal rearrangement. Such a mixture
of products is characteristic of a carbocationic process.
A noticeable feature of the rate data is that the phenyl-
substituted derivative 15b is less reactive than the methyl
compound 15a, by factors of 2.1-4.2 (Table 1). This is contrary
to the behavior observed in unhindered systems, where phenyl
groups accelerate carbocationic reactivity by significant factors
compared to methyl, but is typical of crowded compounds,
where the phenyl is twisted out of conjugation, and destabilizes
the developing carbocationic center due to inductive effects.9
Thus, for compounds RR′2COPNB in 70% acetone at 100 °C
the rate ratio kMe/kPh was 0.005 for R′ ) Me, but was 1.6-5.3
for bulky groups R′. In the case of 15b the adjacent tert-butyl
on the cyclopentadienyl ring would preclude any approach to
coplanarity for the phenyl ring in a developing carbocation.
The high dependence of the reaction rate on the solvent
polarity and the diverse array of products are readily explained
by the process of Scheme 2 with initial ionization to a
carbocation (20) which can undergo direct or allylic substitution
to form 19 or 16a, elimination to form 17, or rearrangement to
21 which gives 18.
was assigned as a cyclopentadiene on the basis of its UV
hexane
max
λ
of 271 nm, as compared to values of 280 and 277 nm,
respectively, for 14a and 15a, and the 1H NMR absorptions for
the vinyl CH3 and the 1,4-vinyl hydrogens at δ 2.07, 5.64, and
5.70, respectively, which are distinctly different from the pattern
for the 1,3-disubstituted cyclopentadienes 14a and 15a. The
vinyl proton at δ 5.64 shows a 2.6 Hz coupling to the other
vinyl proton, and the latter also shows coupling to the vinyl
CH3. The diastereotopic CH2 protons of the CH2CF3 group
differ in chemical shift by only δ 0.003, and this small difference
is consistent with this structure, as opposed to an isomeric
structure resembling 16b, as the different alkyl groups in 16a
which render these protons nonequivalent are rather remote. The
fulvene structure of 17, which is bright yellow in color, follows
hexane
max
from the UV λ
) 246 (ꢀ ) 15 000) and 370 (ꢀ ) 420)
nm, which resembles the published UV spectrum of the parent,8
and also from the other consistent spectral data, including
assignment of the vinyl protons from a H-13C heteronuclear
1
correlated spectrum.
Compounds 14a, 15a, and 19 all show two vinyl tert-butyl
groups, one between δ 1.08 and δ 1.10, and one between δ
1
1.21 and δ 1.24, in the H NMR. Therefore, in 16a the vinyl
tert-butyl at C-2 is that at δ 1.19, and the carbinyl tert-butyl at
C-5 is that of δ 0.97.
An estimate of the reactivity of 15a relative to 1-methylcy-
clopentyl trifluoroacetate (22) is obtained by multiplying the
rate of the corresponding p-nitrobenzoate 22-OPNB of 2.11
× 10-9 s-1 in 80% acetone at 25 °C10ab by the kRO CCF /kROPNB
The structure of the Wagner-Meerwein rearranged product
1
18 follows from the H NMR signals for the carbinyl methyl,
geminal methyls, vinyl methyl, and 1,4-dienyl protons at δ
1.106, 1.114, 2.05, 5.87, and 5.92, respectively, and the
consistent 13C NMR signals. The vinyl methyl and 1,4-dienyl
proton signals are similar to those of 16a, and quite different
from those of the 1,3-disubstituted derivatives 14a and 15a,
2
3
rate factor of 5.9 × 103 (derived for R ) cumyl in MeOH at 25
TFE
1-AdO2CCF3
°C)7e,10c,d,11a and the solvent correction factor k
/
80A
TFE
25
k
of 4507c,11b at 50 °C to give a k
(22) of 5.6 ×
1-AdO2CCF3
10-3 s-1, and a rate ratio, k(22)/k(15a), of 1.6 × 102.
which show the vinyl CH3 at higher field, and a significantly
hexane
max
(9) Tanida, H.; Matsumura, H. J. Am. Chem. Soc. 1973, 95, 1586-1593.
(10) (a) Brown, H. C.; Rao, C. G.; Ravindranathan, M. J. Org. Chem.
1978, 43, 4939-4943. (b) Peters, E. N.; Brown, H. C. J. Am. Chem. Soc.
1975, 97, 2892-2895. (c) Liu, K.-T.; Chen, H.-I.; Chin, C.-P. J. Phys. Org.
Chem. 1991, 4, 463-466. (d) Liu, K.-T.; Chang, L.-W.; Chen, P.-S. J. Org.
Chem. 1992, 57, 4791-4793. (e) Friedrich, E. C.; Taggart, D. B. J. Org.
Chem. 1978, 43, 805-808.
greater spacing between the vinyl protons. The UV λ
)
246 (ꢀ ) 2000) is somewhat shifted from those of 14a, 15a,
16a, and 19, perhaps because of the different substitution pattern
(8) Meuche, D.; Neuenschwander, M.; Schaltegger, H.; Schlunegger, H.
U. HelV. Chim. Acta 1964, 47, 1211-1215.

2374 J. Am. Chem. Soc., Vol. 119, No. 10, 1997
Allen et al.
An estimate of the rates of 15a relative to 9-methyl-9-
fluorenyl trifluoroacetate (23) and 3-methyl-3-indenyl trifluo-
roacetate (24) is obtained by multiplying the rates of 23-ODNB
and 24-ODNB in TFE at 25 °C of 1.76 × 10-4 and 2.0 ×
10-6 calculated from published10e data by the RO2CCF3/ROPNB
rate factor (above) of 5.9 × 103 to give rate constants of 1.04
and 1.2 × 10-2 s-1 for 23 and 24 in TFE at 25 °C, and relative
rates for 23:24:15a of 3.0 × 104:3.5 × 102:1.0. As discussed
below10a an allylic methyl group accelerates the solvolysis of
an allylic substrate by a factor of 2 × 103, and if the two tert-
butyl groups in 15a accelerate the solvolysis by comparable
factors, then the reactivities of 23 and 24 relative to that of a
derivative of 15a without these groups would be correspondingly
greater.
Even though the reactivities of fluorenyl, indenyl, and
cyclopentadienyl substrates leading to carbocationic intermedi-
ates are enormously depressed relative to model compounds in
which antiaromaticity effects are not present, these reactions
can nevertheless be carried out at ambient temperatures at
reasonable rates. It would appear that the lower limits of
reactivity in this series have not yet been reached, and that in
particular the carbocationic reactivity of less highly substituted
derivatives of 15a should be observable.
Experimental Section
Reagents and solvents were the best commercial grade available and
used as supplied except as indicated. For reactions under an inert
atmosphere glassware was flame dried under Ar three times prior to
use. Radial chromatography was carried out using a Chromatotron
from Harrison Research with silica gel coated plates and solvents as
indicated.
2,4-Di-tert-butylcyclopentadienone (13) was prepared by the reported
1
procedure6a and was identified by its H NMR spectrum (CDCl3): δ
1.14 (s, 9, t-Bu), 1.17 (s, 9, t-Bu), 4.98 (d, 1, J ) 1.7 Hz, CdCH),
6.53 (d, 1, J ) 1.7 Hz, CdCH).
1,3-Di-tert-butyl-5-hydroxy-5-methylcyclopenta-1,3-diene (14a).
To freshly prepared ketone 13 (1.0 g, 5.4 mmol) in 60 mL of ether in
a 100 mL three-neck flask under Ar cooled in dry ice/acetone was added
CH3Li (4 mL, 1.5 M in ether, 6.0 mmol) over 5 min, followed by
stirring for 2 h, warming to 25 °C, and stirring overnight. Wet ether
was added followed by H2O and 10 mL of 1 M HCl. The layers were
separated, the H2O layer was washed twice with ether, and the combined
ether layers were extracted by H2O and the brine, dried over CaSO4,
and evaporated to give 1.08 g of crude product, which was chromato-
graphed three times (10:90 EtOAc/hexane, Rf ) 0.25) to give 14a (0.427
g, 2.05 mmol, 38%) as a white solid: mp 84-85 °C; IR (CDCl3) 3592
Thus, the reactivity of the cyclopentadienyl derivative 15a
is appreciably diminished relative to those of 23 and 24.
Moreover, this substrate is even more unreactive compared to
other model compounds. A rate for 3-methyl-3-cyclopentenyl
trifluoroacetate (25) may be derived from the reported10a rate
constant for 25-OPNB in 80% acetone at 25 °C of 1.15 s-1
,
and the kRO CCF /kROPNB and kTFE/k80A rate factors of 5.9 × 103
2
3
TFE
25
and 450 noted above to give a k
(25) of 3.1 × 106 s-1, and
a k(25)/k(15a) of 9.0 × 1010. The rate of 15a would also be
significantly enhanced by the allylic tert-butyl substituent, as
can be noted from the rate factor10a k(CH3CHdCHCMe2OPNB)/
k(CH2dCHCMe2OPNB) of 2.2 × 103 in 80% acetone at 25
°C. Combining these factors gives a predicted rate factor of
2.0 × 1014 for 26/15a in TFE at 25 °C.
1
cm-1 (OH); H NMR (CDCl3) δ 1.08 (s, 9, t-Bu), 1.24 (s, 9, t-Bu),
1.53 (s, 3, CH3), 1.56 (s, 1, OH), 5.48 (d, 1, J ) 2.0 Hz, CH)C), 5.87
(d, 1, J ) 2.0 Hz, CdCH); 13C NMR (CDCl3) δ 23.8, 28.7, 30.8, 31.7,
34.2, 86.1, 123.6, 131.3, 151.7, 161.2; EIMS m/z 208 (M+, 16), 193
(M+ - CH3, 30), 152 (M+ - C4H8, 32), 137 (100), 57 (C4H9+, 93);
HRMS m/z calcd for C14H23O 207.1749, obsd 207.1752; UV λhexane
max
280 nm (ꢀ ) 1900).
1,3-Di-tert-butyl-5-methyl-5-cyclopenta-1,3-dienyl Trifluoroac-
etate (15a). To a solution of alcohol 14a (0.345 g, 1.66 mmol) and
dry pyridine (0.190 mL, 2.35 mmol) in a 25 mL flask with a magnetic
stirrer was added (CF3CO)2O (0.440 mL, 3.11 mmol). There was an
exothermic reaction and precipitation of solid. After stirring for 15
min, ether was added and the mixture was poured onto ice/water and
extracted three times with ether. The combined ether layers were
extracted with NaHCO3 solution and brine, dried over CaSO4, and
evaporated to give 0.366 g of a yellow liquid containing no 14a by 1H
NMR. Chromatography (5:95 EtOAc/hexane, Rf ) 0.62) gave 15a
(0.269 g, 0.885 mmol, 53%) as a pale yellow liquid: IR (CDCl3) δ
Large rate decelerations due to antiaromatic effects have
previously been noted in the fluorenyl5a,10e and indenyl
systems,5b,10e which we have estimated as 103 and 106,
respectively, and the effects in the cyclopentadienyl are even
more dramatic. The 1014 rate depression in TFE at 25 °C for
15a relative to 26 is truly remarkable, and ranks with the largest
of the effects on reactivity due to crowding, strain, and electronic
substituent effects. Together these constitute the four major
structural effects which have dominating influences on organic
reactivity.
In summary the cyclopentadienyl trifluoroacetates 15a and
15b are indicated to undergo solvolysis via cyclopentadienyl
cation intermediates on the basis of their reaction products, the
dependence of their reactivity upon solvent polarity, and the
R-Me/R-Ph rate ratio. The reactivity of 15a is less than that of
the cyclopentenyl analogue 26, by a factor of 1014 in TFE at 25
°C, and this enormous difference may be ascribed to the effects
of antiaromaticity.
1
1778 cm-1 (CdO); H NMR (CDCl3) δ 1.10 (s, 9, t-Bu), 1.21 (s, 9,
t-Bu), 1.74 (s, 3, CH3), 5.75 (d, 1, J ) 2.0 Hz, CdCH), 6.06 (d, 1, J
) 1.9 Hz, CdCH); 13C NMR (CDCl3) 21.8, 28.5, 30.5, 31.6, 34.1,
1
95.3, 114.3 (q, JCF ) 287 Hz, CF3), 124.5, 126.5, 154.9, 155.5 (q,
2JCF ) 42 Hz, COCF3), 156.8; 19F NMR (CDCl3) δ -75.75; UV
hexane
max
λ
277 nm (ꢀ ) 2900); EIMS m/z 304 (M+, 6), 289 (M+ - CH3),
57 (C4H9+, 100); HRMS m/z calcd for C16H23O2F3 304.1650, obsd
304.1636.
Trifluoroethanolysis of 15a. In a preparative reaction 15a (0.170
g, 0.558 mmol) in 25 mL of CF3CH2OH with 2,6-lutidine (0.130 mL,
1.12 mmol) was heated at 46 °C for 250 min. The solution was added
to 12 mL of H2O and extracted five times with pentane, and the pentane
from the combined organic layers was removed by slow distillation.
The crude product was chromatographed with hexane to give five
products, of which the first (trace) in order of elution was unidentified
and the others were identified as 16-19, respectively, in order of
(11) (a) kRO CCF ) 1.13 × 10-3 s-1 and kROPNB ) 1.92 × 10-7 s-1; this
2
3
factor is derived in a more direct fashion than that of 4.5 × 105 reported in
ref 7a, and appears preferable for use. (b) A similar average factor of 420
has been derived from measurements of seven different substrates.10c,d
(12) (a) Cowell, G. W.; Ledwith, A. J. Chem. Soc. B 1967, 695-697.
(b) Brown, H. C.; Dickason, W. C. J. Am. Chem. Soc. 1969, 91, 1226-
1228.
1
elution, in a ratio of 66:10:14:10, as determined from the H NMR of
the crude product.
2,5-Di-tert-butyl-3-methyl-5-cyclopenta-1,3-dienyl 2,2,2-Trifluo-
roethyl Ether (16a): 1H NMR (CDCl3) δ 0.97 (s, 9, t-Bu), 1.19 (s, 9,

Generation of Antiaromatic Cyclopentadienyl Cations
J. Am. Chem. Soc., Vol. 119, No. 10, 1997 2375
t-Bu), 2.07 (d, 3, J ) 1.9 Hz, CH3CdC), 3.50 (q, 2, JHF ) 9.9 Hz,
CH2CF3), 5.62-5.73 (m, 2, 2CH)C); 13C NMR (CDCl3) δ 17.1, 26.1,
16b. The presence of 3% 16b in the product mixture could not be
excluded on the basis of the H NMR. The products were identified
1
2
29.7, 33.2, 36.2, 61.4 (q, JCF ) 33.9 Hz, CH2CF3), 93.5 124.5, (q,
by comparison of their EIMS peaks to those for the corresponding
products from trifluoroethanolysis, as follows: (16a) m/z 241 (M+, 19),
226 (M+ - CH3, 22), 185 (M+ - C4H8, 48), 170 (M+ - C5H11, 100),
152 (90), 57 (C4H9+, 92); (16b) m/z 241 (M+, 13), 226 (4), 185 (49),
170 (100), 152 (64), 57 (55); (17) m/z 192 (M+ + 2, 5), 191 (M+ +1,
42), 190 (M+, 11), 177 (M+ - CH3 +2), 176 (M+ - CH3 +1, 99),
175 (M+ - CH3, 59); 135 (M+ - C4H7, 45), 134 (M+ - C4H8, 100),
133 (M+ - C4H9, 66), 57 (C4H9+, 66); (18) m/z 134, 119, 92 (C2D5-
1JCF ) 278 Hz), 128.6, 134.0, 144.9, 157.7; 19F NMR (CDCl3) δ -75.37
(t, JHF ) 9 Hz); UV λhexane 271 nm (ꢀ ) 930); EIMS m/z 290 (M+, 7),
max
275 (M+ - CH3, 4), 234 (M+ - C4H8, 33), 219 (M+ - C5H11, 73),
178 (22), 57 (C4H9+, 100); HRMS m/z calcd for C16H25OF3 290.1858,
obsd 290.1862.
1,3-Di-tert-butyl-5-methylene-1,3-cyclopentadiene (17): 1H NMR
(CDCl3) δ 1.13 (s, 9, t-Bu), 1.15 (s, 9, t-Bu), 5.61 (d, J ) 1.8 Hz,
C4H), 5.65 (br s, 1, CHH), 5.86 (br s, 1, CHH), 6.13 (t, 1, C2H),
OCMe2+), 57 (t-Bu+); (19) m/z 241 (M+, 16), 226 (M+ - CH3, 26),
+
185 (M+ - C4H8, 49), 170 (M+ - C5H11, 81), 152 (63), 57 (C4H9
100).
,
1
assignments confirmed by the H,13C-heteronuclear coupled spectrum
(2D HMQC); 13C NMR (CDCl3) δ 29.2, 31.96, 32.02, 33.09, 116.6 (d,
1JCH ) 161.8 Hz, C4), 119.6 (t, 1JCH ) 158.1 Hz, CH2), 128.1 (d, J )
1,3-Di-tert-butyl-5-hydroxy-5-phenylcyclopenta-1,3-diene (14b).
As described for 14a the reaction of 13 (0.279 g, 1.45 mmol) with
PhLi (0.9 mL, 1.8 M in cyclohexane/ether, 1.62 mmol) gave 0.290 g
161.8 Hz, C2), 146.6, 151.0, 155.9; UV λhexane 246 (ꢀ ) 15 000), 370
max
(ꢀ ) 420) nm; EIMS m/z 190 (M+, 40), 175 (M+ - CH3, 100), 133
(M+ - C4H9, 87), 119 (M+ - C5H11, 50), 57 (C4H9+, 52); HRMS m/z
calcd for C14H22 190.1722, obsd 190.1725.
1
of an oil which by H NMR contained 50% of the desired product.
Chromatography (5:95 EtOAc/hexane, Rf ) 0.25) gave 14b (0.080 g,
0.03 mmol, 20%) as a pale yellow oil: IR (CDCl3) 3601 cm-1 (OH);
1H NMR (CDCl3) δ 0.98 (s, 9, t-Bu), 1.13 (s, 9, t-Bu), 1.95 (brd s, 1,
OH), 5.62 (d, 1, J ) 1.9 Hz, CHdC), 6.05 (d, 1, J ) 2.1 Hz, CHdC),
7.1-7.4 (m, 5, Ph); 13C NMR (CDCl3) δ 28.7, 30.8, 32.0, 34.4, 89.4,
124.8, 125.2, 126.2, 127.9, 133.0, 140.4, 153.0, 163.8; EIMS m/z 270
3,5-Dimethyl-2-tert-butyl-5-[2′-(2′′,2′′,2′′-trifluoroethoxy)-2′-pro-
pyl]-1,3-cyclopentadiene (18): 1H NMR (CDCl3) δ 1.106 (s, 3, CH3),
1.114 (s, 6, CMe2), 1.178 (s, 9, t-Bu), 2.05 (d, 3, J ) 1.5 Hz, CH3CdC),
3.73 (q, 2, JHF ) 8.7 Hz, CH2CF3), 5.87 (d, 1, J ) 1.4 Hz, CHdC),
5.92 (m, 1, CdCH); 13C NMR (CDCl3) δ 16.4, 17.2, 21.2, 29.7, 29.8,
(M+, 50), 255 (M+ - CH3, 91), 214 (M+ - C4H8, 56), 199 (M+
-
2
1
32.8, 59.2, 60.7 (q, JCF ) 33.7 Hz), 79.1, 124.4 (q, JCF ) 278 Hz),
C5H11, 94), 158 (48) , 57 (C4H9+, 100); HRM S m/z calcd for C19H26O
270.1984, obsd 270.1991.
135.4, 140.5, 140.8, 153.6; 19F NMR (CDCl3) δ -75.1 (t, JHF ) 8.5
hexane
Hz); UV λ
246 nm (ꢀ ) 2000); EIMS m/z 290 (M+, 3), 219 (7),
max
150 (5), 141 (CF3CH2OCMe2+, 100); HRMS m/z calcd for C16H25F3O
290.1858, obsd 290.1847.
1,3-Di-tert-butyl-5-phenyl-5-cyclopenta-1,3-dienyl Trifluoro-
acetate (15b). As described for 15a the reaction of 14b (22 mg, 0.082
mmol) with pyridine (8 µL, 0.1 mmol) and (CF3CO)2O (44 µL, 0.31
mmol) gave after chromatography (10:90 EtOAc/hexane, Rf ) 0.6) 15b
(20 mg, 0.053 mmol, 65%) as a clear liquid: IR (CDCl3) 1789 cm-1
1,3-Di-tert-butyl-5-methyl-5-cyclopenta-1,3-dienyl 2,2,2-Trifluo-
roethyl Ether (19): 1H NMR (CDCl3) δ 1.09 (s, 9, t-Bu), 1.21 (s, 9,
t-Bu), 1.55 (s, 3, Me), 3.50 and 3.51 (ea q, 1, JHF ) 8.9 Hz, CH2CF3),
1
5.35 (bd, 1, J ) 1.6 Hz, CdCH), 6.05 (d, 1, J ) 2.0 Hz, CdCH); 13
C
(CdO); H NMR (CDCl3) δ 0.92 (s, 9, t-Bu), 1.17 (s, 9, t-Bu), 5.78
2
(d, 1, J ) 2.0 Hz, CHdC), 6.24 (d, 1, J ) 2.0 Hz, CHdC), 7.2-7.4
(m, 5, Ph); 13C NMR (CDCl3) δ 28.5, 30.3, 32.2, 34.3, 97.2, 114.3 (q,
1JCF ) 287 Hz), 124.4, 127.2, 127.4, 128.37, 128.44, 135.4, 154.3 (q,
2JCF ) 42 Hz), 155.9, 158.7; 19F NMR (CDCl3) δ -75.3; EIMS m/z
366 (M+, 10), 310 (M+ - C4H8, 40), 57 (C4H9+, 100); HRMS m/z
calcd for C21H25F3O2 366.1807, obsd 366.1818.
NMR (CDCl3) δ 22.7, 28.7, 30.6, 31.9, 33.8, 61.8 (q, JCF ) 34 Hz),
1
90.8, 124.3 (q, JCF ) 278 Hz), 126.8, 127.4, 154.5, 157.9; 19F NMR
(CDCl3) -74.6 (t, J ) 8.9 Hz); UV λhexane 273 nm (ꢀ ) 2000); EIMS
max
m/z 290 (M+, 10), 275 (M+ - CH3, 22), 234 (M+ - C4H8, 43), 219
(M+ - C5H11, 58), 141 (34), 57 (C4H9+, 100); HRMS m/z calcd for
C16H25F3O 290.1858, obsd 290.1860.
Trifluoroethanolysis of 15b. As for 15a a solution of 15b (12 mg,
0.032 mmol) in 0.6 mL of CF3CD2OD was heated to 61 °C for 2 h.
After workup 1.9 mg of material was obtained which by mass spectral
analysis contained unreacted 15b or an isomer (EIMS m/z 366 (M+))
and substitution product derived from 15b (R1 ) CF3CH2) (EIMS m/z
354 (M+), 298 (M+ - C4H8)).
Kinetic Measurements. Kinetics were measured using Perkin-
Elmer Lambda 12 and Varian 210 spectrophotometers using the general
procedures and solvent preparation as reported previously.5 In a typical
procedure solutions of 15a (2 µL, 0.118 M in CH3CN) and 2,6-lutidine
(4 µL, 0.118 M in CH3CN) were added to 1.2 mL of 97% TFE, and
the decrease in absorbance (∆A ) 0.1) was monitored at 300 nm. Stable
end points were observed, with an isobestic point at 335 nm and a
final λmax at 375 nm, primarily due to fulvene 17. In the absence of
buffer the end point was not stable, and the rates were measured using
0.1-0.25 times the concentration of 15a used in the presence of 2,6-
lutidine. For measurements of 15b a solution of the substrate (4 µL,
0.01 M in CH3CN) was added to 1.2 mL of 97% TFE, and the increase
in absorbance was monitored at 272 nm (∆A ) 0.05). For reactions
in HFIP the concentrations of 15b were 2-5 times larger.
In a separate experiment a solution of 15a (3.5 mg, 0.0115 mmol)
and 2,6-lutidine (2.7 µL, 0.023 mmol) in 1 g of CF3CD2OD in an NMR
tube was monitored by 1H NMR and showed a half-life for reaction of
about 47 min, and after 247 min showed no residual 15a. The reaction
mixture was poured into water and extracted with ether which was
dried and evaporated, and the product was chromatographed (5:95
EtOAc/hexane) and analyzed by GC/MS to give peaks with the
following retention times (min), with identification as shown (Vide infra)
based on their MS patterns: 7.9 (unidentified, but probably 16b, as
+
ions at m/z 236 (M+ - C4H8), 221 (M+, - C5H11), and 57 (C4H9
were present), 7.97 (19), 8.03 (17), 8.13 (16a), and 8.75 (18).
)
The alcohol 14a was not observed in any of the product studies,
and was shown to be stable under the reaction conditions.
Ethanolysis of 15a. A solution of 15a (6.4 mg, 0.0211 mmol) in
0.65 mL of EtOH-d6 with 2,6-lutidine (5 µL, 0.043 mmol) in an NMR
1
tube was heated at 100 °C and monitored by H NMR. During the
reaction the ratio of the products formed showed some variation with
time, but after 6.5 h 15a had disappeared and the ratio of products
appeared constant. The reaction mixture was poured into H2O and
pentane, the aqueous layer was extracted five times with pentane, and
the combined organic layers were dried and evaporated, and by analogy
with the product from trifluoroethanolysis of 15a appeared by 1H NMR
to contain the ether 16, fulvene 17, and the ether 19 in a ratio of 42:
40:18. No signals corresponding to those of 18 were detected. Analysis
of the products by GC/MS showed 5% unreacted starting material and
the products 16a, 17, 18, and 19 (R1 ) C2D5) in relative percentages
of 43, 32, <1, and 17%, respectively, in good agreement with the NMR
results. In addition a signal corresponding to 3% of the total, and with
a fragmentation pattern similar to 16a, was observed, which may be
Acknowledgment. Financial support by the Natural Sciences
and Engineering Research Council of Canada is gratefully
acknowledged.
Supporting Information Available: 1H NMR spectra (8
pages). See any current masthead page for ordering and Internet
access instructions.
JA962494Z