Mechanisms of solvent- and base-promoted imine-forming elimination reactions

Article abstract of DOI:10.1021/ja9624681

Solvolysis of 9-(N-chloro-N-methylamino)fluorene (1-Cl) in 25 vol % acetonitrile in water gives the imine fluorenylidenemethylamine (3) as the sole product. The kinetic deuterium isotope effect was measured with the deuterated analogue (9-²H)-9

Full text of DOI:10.1021/ja9624681

1224
J. Am. Chem. Soc. 1997, 119, 1224-1229
Mechanisms of Solvent- and Base-Promoted Imine-Forming
Elimination Reactions
Qingshui Meng and Alf Thibblin*
Contribution from the Institute of Chemistry, UniVersity of Uppsala,
P.O. Box 531, S-751 21 Uppsala, Sweden
ReceiVed July 17, 1996X
Abstract: Solvolysis of 9-(N-chloro-N-methylamino)fluorene (1-Cl) in 25 vol % acetonitrile in water gives the imine
fluorenylidenemethylamine (3) as the sole product. The kinetic deuterium isotope effect was measured with the
deuterated analogue (9-2H)-9-(N-chloro-N-methylamino)fluorene as kH/kD ) 4.8 ( 0.2 (50 °C) without base and
kH/kD ) 6.7 ( 0.2 (25 °C) with hydroxide anion. The solvent- and base-promoted reactions of 1-Cl are concluded
to be of E2 type. The corresponding substrates 9-(N-4-Y-benzenesulfonyl-N-methylamino)fluorene (2-Y, Y ) OMe,
Me, or Br), with very poor leaving groups, show reversible E1cB reactions with added bases. The strongly activated
substrate 9-(N-4-nitrobenzenesulfonyl-N-methylamino)fluorene (2-NO2) does not give any elimination; it exclusively
undergoes intramolecular nucleophilic aromatic substitution involving rate-limiting hydron transfer.
Introduction
Scheme 1
We are interested in the borderline between stepwise and one-
step alkene-forming elimination reactions and have discussed
this subject in a number of recent papers.1-9 Such questions
as what can induce a change in mechanism from a stepwise
route involving an intermediate of carbocationic or carbanionic
type to a one-step, concerted process have been addressed.
We have now extended the study to imine-forming elimina-
tion reactions. Such reactions are less complex in the sense
that they generally provide the elimination product exclusively
which can facilitate kinetic studies. However, this can be a
disadvantage in mechanistic studies because branching from a
common intermediate can often give valuable mechanistic
information.3,10
In the present paper we report results of studies on elimination
reactions that form carbon-nitrogen double bonds from sub-
strates which have the â-hydrogen substantially activated. In
the corresponding alkene-forming elimination reactions of
fluorenyl-activated substrates, depending on the substrate struc-
ture and the leaving group, E1cBR, E1cBI, and E2 reactions are
possible in addition to E1 elimination through an ion-pair
intermediate. Moreover, the first conclusive evidence for the
solvent-promoted E2 elimination mechanism was reported
recently for secondary halides in a highly aqueous solvent.6-9
What are the possible mechanisms of imine-forming elimination
reactions? A review by Hofmann, Bartsch, and Cho discusses
one-step, concerted base-promoted imine-forming reactions,11
but there seems to be no report on stepwise carbanionic imine-
forming reactions in the literature.
leaving groups the reaction occurs through the expulsion of the
leaving group from a reversibly formed carbanion intermediate
(E1cBR). A change to a more efficient leaving group changes
the mechanism to E1cBI or to E2. With the efficient chloride
ion leaving group, the very weakly basic solvent (water) is able
to abstract a hydron from the substrate giving water-promoted
elimination.
This report also includes a study of the base-catalyzed reaction
of 2-NO2 which does not give elimination, but reacts by
intramolecular nucleophilic aromatic substitution by a reaction
with rate-limiting hydron transfer.
Results
We now report results for solvent- and base-promoted imine-
forming elimination reactions which show that with very poor
The solvolysis of 9-(N-chloro-N-methylamino)fluorene (1-
Cl) in 25 vol % acetonitrile in water at 50 °C yields exclusively
the elimination product fluorenylidenemethylamine (3) (Scheme
1). The reaction in the aqueous solvent as well as in methanol
is catalyzed by general bases. The kinetics was studied by a
sampling high-performance liquid chromatography (HPLC)
procedure or by following the decrease in absorbance at 275
nm by UV spectrophotometry. The ester 9-(N-benzoyloxy-N-
methylamino)fluorene (1-OCOPh) is less reactive. Its reaction
with bases in methanol affording the imine 3 as the main product
(about 99% with the amines and about 70% with methoxide
ion) has been studied by the same techniques. The rate constants
X Abstract published in AdVance ACS Abstracts, January 15, 1997.
(1) Thibblin, A. J. Am. Chem. Soc. 1987, 109, 2071.
(2) Thibblin, A. J. Am. Chem. Soc. 1988, 110, 4582.
(3) Thibblin, A. Chem. Soc. ReV. 1993, 22, 27 and references therein.
(4) Thibblin, A.; Sidhu, H. J. Chem. Soc., Perkin Trans. 2 1994, 1423.
(5) Thibblin, A.; Sidhu, H. J. Phys. Org. Chem. 1994, 7, 578.
(6) Meng, Q.; Thibblin, A. J. Am. Chem. Soc. 1995, 117, 1839.
(7) Meng, Q.; Thibblin, A. J. Am. Chem. Soc. 1995, 117, 9399.
(8) Meng, Q.; Thibblin, A. J. Chem. Soc., Chem. Commun. 1996, 345.
(9) Meng, Q.; Gogoll, A.; Thibblin, A. J. Am. Chem. Soc. 1996, 119,
1217.
(10) Thibblin, A.; Ahlberg, P. Chem. Soc. ReV. 1989, 18, 209.
(11) Hoffman, R. V.; Bartsch, R. A.; Cho, B. R. Acc. Chem. Res. 1989,
22, 211.
S0002-7863(96)02468-7 CCC: $14.00 © 1997 American Chemical Society

Imine-Forming Elimination Reactions
J. Am. Chem. Soc., Vol. 119, No. 6, 1997 1225
Table 1. Rate Constants for the Reactions of 1-X and 2-Y
substratea
base
base conc, M
106kobs, s-1 k
106k, M-1 s-1 k
25 vol % Acetonitrile in Water
1-Cl
1-Cl
1-Cl
2-Br
2-Me
2-OMe
solventb
solventb,c
HO- e
402 (84.5)l
123 (24.7)l
4800
1140
343
0.0025
0.75
0.75
1.92 (0.287)l × 106
HO- e,j
HO- e,j
HO- e,j
1530
457
423
0.75
317
Methanol
0.00137
0.0025-0.01
0.00497-0.02024
0.02-0.10
0.10-0.50
0.00137
0.0575-0.115
0.50
0.20-0.50
1.37
1.37
1-Cl
1-Cl
1-Cl
1-Cl
MeO- e
4100
3.00 (0.468)l × 106
HFIPe,g
3.68 × 105
3.01 × 105
5.60 × 104
quinuclidinee
DABCOe,h
HMTAe,i
MeO- e
1-Cl
1810
1-OCOPh
1-OCOPh
1-OCOPh
1-OCOPh
2-Br
243
1.22 (0.234)l × 105
quinuclidinee
DABCOe,h
HMTAe,i
MeO- f
1.39 × 104
1010
2020 (326)l
46.7
1160
280
228
745
847 (841)l
204 (204)l
166
2-Me
MeO- f
2-OMe
2-NO2
2-NO2
2-NO2
MeO- f
1.37
0.0137
0.0023-0.0092
0.05-0.20
MeO- e
5.44 (0.942)l × 104
2.39 × 104
986
quinuclidined
DABCOd,h
a Substrate concentration 0.01-0.1 mM. b 50 °C. c In the presence of 1 mM HClO4. d 70 °C. e 25 °C. f 35 °C. g (CF3)2CHO-/(CF3)2CHOH ) 1.
h Diazabicyclo[2.2.2]octane. i Hexamethylenetetramine. j The ionic strength was maintained constant (0.75 M) with sodium perchlorate. k Estimated
maximum error is (5%. l Rate constant with the corresponding 9-(2H) compound.
Table 2. Kinetic Deuterium Isotope Effects for the Reactions of
1-X and 2-Y
substratea
solvent
base
kH/kD
1-Cl
1-Cl
1-Cl
1-OCOPh
1-OCOPh
2-Br
2-Me
2-NO2
25% MeCNd
25% MeCNd
MeOH
MeOH
MeOH
MeOH
MeOH
MeOH
solventb,c
HO- e
5.0 ( 0.3
6.7 ( 0.3
6.4 ( 0.3
5.2 ( 0.3
6.2 ( 0.3
1.0 ( 0.1
1.0 ( 0.1
5.8 ( 0.3
MeO- e
MeO- e
DABCOe,g
MeO- f
MeO- f
MeO- e
a Substrate concentration 0.01-0.1 mM. b 50 °C. c In the presence
of 1 mM HClO4; the isotope effect without acid addition is 4.8 ( 0.2.
d By volume in water. e 25 °C. f 35 °C. g Diazabicyclo[2.2.2]octane.
for the reactions are shown in Table 1 which also includes the
rate constants for the reactions of the (9-2H)-substituted
analogues. The kinetic deuterium isotope effects are collected
in Table 2. Attempts to prepare the bromide and the arylsul-
fonyloxy derivatives have not been successful, probably owing
to the very high reactivity of these compounds.
Figure 1. The rate constants for the imine-forming reactions as a
function of methoxide ion concentration in methanol at 35 °C: 2-Br
(9) and 2-Me (b).
The byproduct (ca. 30%) obtained from the methoxide-
promoted reaction with 1-OCOPh is probably the hydroxy-
lamine derivative formed by ester cleavage. Consistently, the
UV spectrum of this product is very similar to that of the ester.
The product is not stable but reacts to give another product with
a quite different UV spectrum.
The much slower reactions of 2-OMe, 2-Me, and 2-Br to
give the imine 3 as the sole product (Scheme 1) were studied
with strong base, methoxide, or hydroxide anion (Table 1). The
reactions exhibit apparent third-order kinetics, first-order in
substrate and second-order in base, as shown in Figures 1 and
2. However, this behavior may be a medium effect because it
is difficult mechanistically to account for third-order kinetics
(see Discussion). 1H NMR analysis showed complete incor-
poration of 1H in the 9-position of the deuterated substrate d-2-
Me after one half-life with methoxide ion in methanol. Thus,
these reactions are accompanied by complete deuterium-protium
exchange with the solvent and they show no kinetic deuterium
isotope effects.
Figure 2. The rate constants for the imine-forming reaction of 2-Br
as a function of hydroxide ion concentration in 25 vol % acetonitrile
in water at 25 °C; ionic strength maintained constant (0.75 M) with
sodium perchlorate.
The compound 2-NO2 does not undergo elimination with
strong bases; it exclusively yields 4 (Scheme 2). This intramo-
lecular nucleophilic aromatic substitution reaction is general base

1226 J. Am. Chem. Soc., Vol. 119, No. 6, 1997
Meng and Thibblin
Scheme 2
Figure 3. Brønsted plots for the reactions of 1-Cl (b) and 1-OCOPh
(9) with tertiary amines in methanol at 25 °C. The slopes are â )
0.49 and â ) 0.53, respectively. The pKa values refer to water.20 The
rate constants for hexamethylenetetramine (HMTA) and diazobicyclo-
[2.2.2]octane (DABCO) have been statistically corrected.
catalyzed. The measured rate constants and kinetic deuterium
isotope effect are presented in Tables 1 and 2, respectively.
Discussion
Solvent- and Base-Promoted Elimination. The elimination
reactions of 1-Cl and 1-OCOPh are catalyzed by general bases.
The Brønsted parameters of â ) 0.49 and 0.53 for catalysis by
tertiary amines in methanol (Figure 3) indicate substantial
amounts of hydron transfer in the rate-limiting transition state.
Consistently, the kinetic deuterium isotope effects are large
(Table 2). These results strongly indicate E2 and E1cBI
mechanisms since very small â values and small kinetic
deuterium isotope effects are expected for a mechanism in which
a nitrenium ion or ion pair is dehydronated in the rate-limiting
step.
Figure 4. More O’Ferrall-Jencks diagram for imine-forming elimina-
tion.12 The effect of change in leaving group from chloride ion to the
less efficient benzoate ion is a raising of the energy of the upper edge
of the diagram. The result is a shift to the right and an increase in â as
indicated.
The base-promoted alkene-forming dehydrochlorination of
9-(1-chloroethyl)fluorene has been concluded to be of E2
type.6,7,9 A nitrogen-chlorine bond is weaker than a bond
between carbon and chlorine. Thus, 1-Cl should undergo a
facile cleavage of the nitrogen-chlorine bond, and we suggest
that this reaction also occurs by an E2 mechanism. Consistently,
the imine-forming reaction of 1-Cl with hydroxide anion is about
24 times faster than the corresponding alkene-forming reaction.
The methoxide-promoted imine-forming elimination from N-
chloro-N-methylbenzylamine proceeds 104 times faster than that
from the corresponding alkene-forming elimination from 2-chloro-
1-phenylpropane under comparable conditions.11 The much
larger rate ratio for the benzyl substrates is due to a large amount
of N-Cl bond cleavage in the E1-like E2 transition state of the
imine-forming reaction; the alkene-forming reaction has a more
E1cB-like transition state. In contrast, the imine- and alkene-
forming reactions of the fluorene derivatives both have E1cB-
like E2 mechanisms with relatively small amounts of cleavage
of the bond to the chlorine in the transition states.
transition state of â-elimination reactions.12 The effect of a less
good leaving group is to increase the energy level at the upper
edge of the diagram. This results in movement of the transition
state uphill toward the product and downhill perpendicular to
the original reaction coordinate. This corresponds to a shift to
the right and an increase in â as indicated by the arrows in
Figure 4. The small increase in â from 0.49 for 1-Cl to 0.53
for 1-OCOPh is in accord with this model. However, it cannot
be excluded that the mechanism has changed from E2 to E1cBI.
The most significant result with 1-Cl is the spontaneous
elimination reaction observed in the absence of added base. The
substrate should have a pKa not far from zero since it has been
found that the chloro substituent has a very large acidifying
effect; e.g., the pKa of N-chlorodimethylamine is 0.47.13 Thus,
in neutral solution the compound exists in its unhydronated form.
Strong support for a solvent-promoted E2 mechanism for 1-Cl
(12) (a) More O’Ferrall, R. A. J. Chem. Soc. B 1970, 274. (b) Jencks,
W. P. Chem. ReV. 1972, 72, 705. (c) Jencks, D. A.; Jencks, W. P. J. Am.
Chem. Soc. 1977, 99, 7948. (d) Gandler, J. R.; Jencks, W. P. J. Am. Chem.
Soc. 1982, 104, 1937. (e) Lowry, T. H.; Richardson, K. S. In Mechanism
and Theory in Organic Chemistry; Harper & Row: New York, 1987; p
374.
The substrate 1-OCOPh reacts, as expected owing to the less
efficient leaving group, more slowly with bases to give the imine
(about 20-40 times, Table 1). These reactions may be classified
also as E2 reactions as will be shown in the following. More
O’Ferrall-Jencks diagrams of the type shown in Figure 4 have
been used frequently for mapping the characteristics of the
(13) Antelo, J. M.; Arce, F.; Franco, J.; Sanchez, M.; Varela, A. Bull.
Soc. Chim. Belg. 1989, 98, 85.

Imine-Forming Elimination Reactions
J. Am. Chem. Soc., Vol. 119, No. 6, 1997 1227
Figure 5. Brønsted plot (using the second-order rate constants reported
in Table 1) for the imine-forming reaction of the para-substituted
benzenesulfonamides 2-Br, 2-Me, and 2-OMe with 0.75 M NaOH in
25 vol % acetonitrile in water and 1.37 M NaOMe in methanol,
respectively. The pKa values refer to water.20
Figure 6. Hammett plot (using the second-order rate constants reported
in Table 1) for the imine-forming reactions of the para-substituted
benzenesulfonamides 2-Br, 2-Me, and 2-OMe with 0.75 M NaOH in
25 vol % acetonitrile in water and 1.37 M NaOMe in methanol.
Sulfonamides are among the most stable of the nitrogen
protecting groups and drastic deprotection conditions are
generally required. For example, arylsulfonamides are cleaved
by sodium in liquid ammonia or by strong acid at high
temperature.14,15 However, the presence of an acidic hydron
on the â-carbon atom can result in a relatively facile cleavage
of the nitrogen-sulfur bond through an imine-forming elimina-
tion reaction occurring via the reversibly formed carbanion.
Intramolecular Nucleophilic Aromatic Substitution. The
substrate 2-NO2 is more activated and is expected to give imine
product faster than the other three aryl-substituted derivatives.
However, no elimination product is formed; the sole product is
compound 4 (Scheme 2). The large isotope effect of kH/kD )
5.8 (Table 2) shows that the transfer of the hydron occurs in
the rate-limiting transition state. Consistently, the Brønsted
parameter obtained with quinuclidine and DABCO is large, â
) 0.50. Possible mechanisms are shown in Scheme 2 (R- may
be an intermediate or a transition state). The simplest mech-
anism to account for the results is a concerted formation of 4
directly from 2-NO2. Another, more likely, mechanism is the
formation of the carbanion R- in a rate-limiting step followed
by a concerted or stepwise reaction to give the product.
Alternatively, the anion R′- is formed directly from the substrate
followed by fast formation of the product 4. This reaction route
may be facilitated by participation of the aromatic ring in the
ionization.
is the large kinetic isotope effect measured with the 9-deuterated
analogue of kH/kD ) 5.0 ( 0.2 (Table 2), which, using the same
arguments as above, is not consistent with a stepwise route
through an ion or through ion pairs. The reaction is, as expected,
faster than that of 9-(1-chloroethyl)fluorene, which has been
concluded to be of the solvent-promoted E2 type.
Generally, it is possible in base-promoted alkene-forming
elimination reactions to change the mechanism from a concerted
E2 or an irreversible carbanion mechanism, E1cBI, to a
reversible mechanism, E1cBR, by changing to a less efficient
leaving group. We have used the very poor arylsulfinate leaving
groups in this study to ensure a shift in mechanism. Thus, the
reactions of 2-Me and 2-Br with methoxide anion in methanol
do not show any kinetic deuterium isotope effects and they show
complete exchange of the 9-deuterium with protium from the
solvent, which strongly indicate the E1cBR mechanism.
The kinetics of the reactions of the substrates with the very
poor leaving groups, arylsulfinate, seem to be second-order in
methoxide-ion concentration (Figure 1). We also see this effect
with hydroxide anion at constant ionic strength (Figure 2) which
indicates that the effect of methoxide-ion concentration is not
an effect of ion-pairing of the base. However, it is diffcult to
see how the base could assist the departure of the leaving group.
Therefore, the most likely explanation is medium effects.
A large âlg is expected for the E1cBR mechanism. The
reaction rates of the arylsulfonyl derivatives show an unusually
large sensitivity to the pKa of the leaving group, with a Brønsted
slope of âlg ) -4.5 in the aqueous solvent and -5.6 in methanol
(Figure 5). The corresponding Hammett parameters for the
effect of para substitution in the aromatic ring of the leaving
group have modest slopes of Flg ) 1.1 and 1.4, respectively
(Figure 6), indicating some negative charge on the departing
group in the rate-limiting transition state.
An unusually large sensitivity to the pKa of the leaving group
does not explain reasonably why the chloride 1-Cl reacts so
much more quickly than 2-Me; a rate ratio of about 105 can be
estimated from the data of Table 1. This large ratio could be
ascribed to the difficulty in breaking a nitrogen-sulfur bond.
A nitrogen-chlorine bond is much easier to cleave. The barrier
for departure of the chloride leaving group from the putative
carbanion intermediate should be extremely small. In fact, we
believe that there is no barrier for departure of the chloride anion
from the carbanion. This corresponds to an uncoupled concerted
E2 reaction. However, the hydron transfer and the cleavage of
the nitrogen-chlorine bond should be coupled in the E1cB-
like E2 transition state (Vide supra).
Experimental Section
General. The NMR spectra were recorded with a Varian 400
spectrometer: 1H at 400 and 13C at 100.5 MHz using the solvent,
1
chloroform-d1 (7.26 ppm, H, 77.0 ppm, 13C), as an internal standard.
Infrared spectra were measured with a Perkin-Elmer 1600 FT-IR
spectrometer. The high-performance liquid chromatography analyses
were carried out with a Hewlett-Packard 1090 liquid chromatograph
equipped with a diode-array detector on a Asahipak (5 µm, 4 × 125
mm) reversed-phase column. The mobile phase was a solution of
acetonitrile in water. The reactions were studied at constant temperature
in a HETO 01 PT 623 thermostat bath. The UV spectrophotometry
was performed with a Kontron Uvicon 930 spectrophotometer equipped
with an automatic cell changer kept at constant temperature with water
from the thermostat bath. The pH was measured using a Radiometer
PHM82 pH meter with an Ingold micro glass electrode. Elemental
analyses were performed by Analytische Laboratorien, Lindlar, Ger-
many.
(14) Parsons, A. F.; Pettifer, R. M. Tetrahedron Lett. 1996, 37, 1667
and references therein.
(15) Backes, B. J.; Ellman, J. A. J. Am. Chem. Soc. 1994, 116, 11171.

1228 J. Am. Chem. Soc., Vol. 119, No. 6, 1997
Meng and Thibblin
Merck silica gel 60 (249-400 mesh) was used for flash chroma-
tography. Methylene chloride and toluene were distilled under nitrogen
from calcium hydride. Methanol and acetonitrile were of HPLC grade.
Chloroform was dried over 4 Å molecular sieves. Diazabicyclo[2.2.2]-
octane (DABCO) was recrystallized twice from hexane. Quinuclidine
was sublimed twice at reduced pressure. All other reagents were of
reagent grade and were used without further purification. A stock
solution of NaOMe was prepared by adding freshly cut pieces of pure
sodium to dry methanol. The concentration was determined by titration
of aliquots of this stock solution with 1 M HCl. All the deuterated
compounds have at least >98.5 atom % 2H in the 9-position as
with saturated sodium carbonate solution and with water, followed by
drying over sodium sulfate. The solvent was removed to give a yellow
oil, which was purified by flash chromatography with 5% ethyl acetate
in pentane as eluent. After removal of solvent, the residue was
recrystallized from diethyl ether-pentane (1:2) giving a white solid:
mp 112-113 °C; 1H NMR δ 8.07-7.43 (m, 11 H), 7.35 (dt, J ) 7.5,
1.2 Hz, 2 H), 5.60 (s, 1 H), 2.50 (s, 3 H); 13C NMR δ 164.97, 141.40,
141.00, 133.09, 129.50, 129.40, 128.85, 128.56, 127.44, 126.58, 119.96,
71.59, 38.96.
9-(N-p-Toluenesulfonyl-N-methylamino)fluorene (2-Me). A solu-
tion of fluorenylmethylamine in dichloromethane (5 mL, 0.8 mmol)
was cooled to 0 °C and p-toluenesulfonic anhydride (0.261 g, 0.8 mmol)
was added. Then, triethylamine (0.08 g, 0.8 mmol) in 2 mL of
dichloromethane was added to the reaction mixture. The mixture was
stirred for 1 h. It was washed with water and the water phase was
extracted with dichloromethane. The combined organic phases were
washed with water and brine, dried over magnesium sulfate, and
concentrated. Flash chromatography with 50% pentane-acetate gave
1
measured by H NMR analysis.
The deuterated compounds (9-2H)-9-(N-chloro-N-methylamino)-
fluorene (d-1-Cl), (9-2H)-9-(N-p-toluenesulfonyl-N-methylamino)fluo-
rene (d-2-Me), (9-2H)-9-(N-benzoyloxy-N-methylamino)fluorene (d-
1-OCOPh), (9-2H)-9-(N-4-bromobenzenesulfonyl-N-methylamino)-
fluorene (d-2-Br), and (9-2H)-9-(N-4-nitrobenzenesulfonyl-N-methy-
lamino)fluorene (d-2-NO2) were prepared from (9-2H)-fluorenylmethy-
lamine as described for the respective protium compound.
1
a pure white solid: mp 162-163 °C; H NMR δ 7.94 (m, 2 H), 7.66
(dt, J ) 7.5, 0.9 Hz, 2 H), 7.42 (m, 2 H), 7.37 (m, 2 H), 7.22 (td, J )
7.5, 1.1 Hz, 2 H), 7.11 (m, 2 H), 6.00 (s, 1 H), 2.52 (s, 3 H), 2.29 (s,
3 H); 13C NMR δ 143.53, 141.40, 140.64, 137.81, 129.95, 128.86,
127.60, 127.32, 125.16, 119.99, 62.55, 29.25, 21.59.
9-(N-4-Methoxybenzenesulfonyl-N-methylamino)fluorene (2-
OMe) was synthesized as described above with fluorenylmethylamine
and 4-methoxybenzenesulfonyl chloride. Recrystallization from etha-
nol-chloroform-pentane (2:1:2) gave pure material: mp 195-196 °C;
1H NMR δ 7.98 (m, 2 H), 7.67 (dt, J ) 7.6, 0.9 Hz, 2 H), 7.37 (tdd,
J ) 7.5, 1.1, 0.6 Hz, 2 H), 7.23 (td, J ) 7.5, 1.1 Hz, 2 H), 7.15 (dq,
J ) 7.6, 0.9 Hz, 2 H), 7.09 (m, 2 H), 5.99 (s, 1 H), 3.94 (s, 3 H), 2.28
(s, 3 H).
9-(N-4-Bromobenzenesulfonyl-N-methylamino)fluorene (2-Br)
was synthesized as described above with fluorenylmethylamine and
4-bromobenzenesulfonyl chloride. Recrystallization from ethanol-
chloroform-pentane (2:1:2) gave pure material: mp 155-156 °C; 1H
NMR δ 7.91 (m, 2 H), 7.77 (m, 2 H), 7.67 (dt, J ) 7.6, 0.9 Hz, 2 H),
7.38 (tdd, J ) 7.5, 1.1, 0.6 Hz, 2 H), 7.24 (td, J ) 7.5, 1.1 Hz, 2 H),
7.11 (dq, J ) 7.6, 0.9 Hz, 2 H), 5.99 (s, 1 H), 2.29 (s, 3 H); 13C NMR
δ 141.05, 140.66, 139.84, 132.66, 129.05, 128.76, 127.74, 127.67,
125.00, 120.13, 62.64, 29.30.
Fluorenylidenemethylamine (3) was prepared by a slight modifica-
tion of a previously published method.16 Methylamine gas was bubbled
through a solution of 9-fluorenone (5.4 g, 15 mmol) dissolved in toluene
(40 mL) at 0 °C for 20 min. Titanium tetrachloride (30 mmol) in
dichloromethane (15 mL) was added dropwise from a syringe. After
addition, methylamine gas was bubbled through the reaction mixture
again until the red color had disappeared. The reaction mixture was
warmed to room temperature and stirred for 2 h. Then, it was poured
into water and extracted with ethyl acetate, followed by washing with
water and brine, and drying over sodium sulfate. The solvent was
removed and the residue recrystallized from ether-hexane (1:2) which
gave a yellow solid (5.2 g, 90%): mp 51-52 °C (lit.17 110-111 °C);
1H NMR δ 7.94 (m, 1 H), 7.77 (dt, J ) 7.5, 1.0 Hz, 1 H), 7.68 (ddd,
J ) 7.5, 1.2, 0.8 Hz, 1 H), 7.59 (dt, J ) 7.5, 1.0 Hz, 1 H), 7.44 (td, J
) 7.5, 1.2 Hz, 1 H), 7.40 (td, J ) 7.5, 1.2 Hz, 1 H), 7.30 (tt, J ) 7.5,
1.2 Hz, 2 H), 3.98 (s, 3 H); 13C NMR δ 164.68, 143.48, 140.97, 138.30,
131.92, 131.16, 130.73, 128.35, 127.90, 127.74, 122.14, 120.30, 119.34,
41.17.
Fluorenylmethylamine.18 Sodium cyanoborohydride (NaBH3CN,
0.4 g, 6.4 mmol) and acetic acid (1 mL) were added to fluorenyliden-
emethylamine (0.6 g, 3.2 mmol) dissolved in methanol (25 mL). The
mixture was heated at 40 °C for 1 h. Methanol was removed and 2 M
sodium hydroxide (10 mL) was added to the residue, followed by
extraction with ethyl acetate, washing with water and brine, and drying
over magnesium sulfate. After evaporation of the solvent, the residue
was dissolved in dry dichloromethane (20 mL) and stored in the
refrigerator under nitrogen. The compound was used to prepare 1-X
and 2-Y; the total yields of these two-step syntheses were 50-80%
based on fluorenylidenemethylamine.
(9-2H)-Fluorenylmethylamine was prepared from fluorenyliden-
emethylamine by reduction with sodium borodeuteride in MeOD but
otherwise as described above. The washing of the extract was done
with D2O.
9-(N-Chloro-N-methylamino)fluorene (1-Cl). The solution of
fluorenylmethylamine in dichloromethane from the previous step (5
mL, 0.8 mmol) was cooled to 0 °C and N-chlorosuccinimide (0.107 g,
0.8 mmol) was added. The mixture was stirred at 0 °C for 1 h. The
solvent was removed and the residue was extracted with hexane twice.
Recrystallization from ethanol-pentane (1:1) gave pure material: mp
80-81 °C (decomposes to an orange solid of the corresponding imine);
1H NMR δ 7.84 (m, 2 H), 7.69 (dt, J ) 7.5, 1.0 Hz, 2 H), 7.43 (tdd,
J ) 7.5, 1.2, 0.6 Hz, 2 H), 7.33 (td, J ) 7.5, 1.2 Hz, 2 H), 5.37 (s,1
H), 2.53 (s, 3 H).
9-(N-4-Nitrobenzenesulfonyl-N-methylamino)fluorene (2-NO2)
was synthesized as described above from fluorenylmethylamine and
4-nitrobenzenesulfonyl chloride. Recrystallization from ethanol-
chloroform-pentane (2:1:2) gave pure material: mp 182-183 °C; 1H
NMR δ 8.49 (m, 2 H), 8.24 (m, 2 H), 7.69 (dt, J ) 7.5, 1.0 Hz, 2 H),
7.40 (tdd, J ) 7.5, 1.1, 0.9 Hz, 2 H), 7.24 (td, J ) 7.5, 1.1 Hz, 2 H),
7.09 (dq, J ) 7.5, 1.0 Hz, 2 H), 6.02 (s, 1 H), 2.35 (s, 3 H); 13C NMR
δ 150.09, 146.62, 140.71, 140.65, 129.27, 128.36, 127.86, 124.79,
124.68, 120.29, 62.82, 29.42. Anal. Calcd for C20H16N2O4S: C, 63.14;
H, 4.24; N, 7.37; S, 8.41. Found: C, 62.89; H, 4.22; N, 7.36; S, 8.46.
9-(4-Nitrophenyl-9-N-methylamino)fluorene (4). 9-(N-4-Nitroben-
zenesulfonyl-N-methylamino)fluorene (2-NO2, 50 mg) was dissolved
in a solution of 1 M NaOMe (40 mL) in methanol. The mixture was
stirred at room temperature for 1 h and the solvent was removed. The
residue was dissolved in ether and the solution was washed with water
and brine and dried over magnesium sulfate. Evaporation of the ether
gave a pure yellow solid material: mp 159-160 °C; 1H NMR δ 8.07-
7.20 (m, 12 H), 2.06 (s, 3 H), 1.91 (s, 1 H); 13C NMR: δ 152.33,
148.31, 146.99, 140.56, 128.71, 128.11, 127.27, 124.53, 123.49, 120.24,
73.54, 30.12; IR (CDCl3) 3326, 3065, 2949, 2797, 1604, 1594, 1517,
1489, 1474, 1449, 1349, 1109, 852 cm-1
.
Anal. Calcd for
C20H16N2O2: C,75.93; H, 5.10; N, 8.85. Found: C, 75.77; H, 5.24;
N, 8.75.
9-(N-Benzoyloxy-N-methylamino)fluorene (1-OCOPh). To the
solution of fluorenylmethylamine in chloroform (10 mL, 1.07 mmol)
were added dibenzoyl peroxide (0.24 g, 1.07 mmol) and potassium
carbonate (0.52 g).19 The mixture was refluxed for 2.5 h at 65 °C.
The mixture was filtered and the chloroform phase was washed twice
Kinetics and Product Studies. The reaction solution was prepared
by mixing acetonitrile with water at room temperature, ca. 22 °C. It
was transferred into several 2-mL HPLC flasks, which were sealed
with gas-tight PTFE septa and placed in an aluminum block in the
water thermostat bath. The concentration of the substrate in the reaction
solution was usually about 0.01-0.1 mM. At appropriate intervals,
(16) Weingarten, H.; Chupp, J. P.; White, W. A. J. Org. Chem. 1967,
32, 3246.
(17) Reddelien, G. Berichte 1921, 54, 3121.
(18) Bordwell, F. G.; Lynch, T. Y. J. Am. Chem. Soc. 1989, 111, 7558.
(19) Psiore, M.; Zinner, G. Synthesis 1984, 217.
(20) Jencks, W. P.; Regenstein, J. Handbook of Biochemistry and
Molecular Biology, 3rd ed.; Fasman, G. D., Ed.; CRC Press: Cleveland,
1976.

Imine-Forming Elimination Reactions
J. Am. Chem. Soc., Vol. 119, No. 6, 1997 1229
samples were analyzed using the HPLC apparatus. For the reaction
of 2-NO2 with sodium methoxide, the samples were quenched with an
aqueous solution of acetic acid (1 M) before analysis. The rate
constants for the disappearance of the substrates were calculated from
plots of substrate-peak area vs time by means of a nonlinear regression
computer program. Very good pseudo-first-order behavior was seen
for all of the reactions studied.
The estimated errors are considered as maximum errors derived from
maximum systematic errors and random errors.
Hydrogen-Exchange Experiments. The deuterated substrate d-2-
Me (3 mg) was dissolved in a solution of NaOMe (10 mL, 1.37 M) in
MeOH. The mixture was stirred at room temperature for 2 h (ca. 1
half-life). The reaction solution was quenched with 15 mL of 2 M
sulfuric acid. The mixture was extracted with diethyl ether, washed
with water and brine, and dried over sodium sulfate. Ether was removed
and the residue was dissolved in chloroform-d1. 1H NMR analysis
showed that the 2H in 9-position of the starting material (d-2-Me) had
been exchanged completely.
The fast reactions of 1-X and 2-Y with strong bases were followed
for at least 10 half-lives by monitoring the decrease in absorbance at
275 nm by using thermostated 3-mL quartz cells as reaction vessels.
After complete reaction, the reaction mixture was quenched with acetic
acid (1 M aqueous solution). Analysis by HPLC showed that
fluorenone is the only product. The relative response factors of 2-Me,
1-OCOPh, and fluorenone were determined by weighing both of them
into a flask, dissolving in acetonitrile, and analyzing by HPLC.
Acknowledgment. We thank the Swedish Natural Science
Research Council for supporting this work.
JA9624681