Cyclopropanediamines. 6. Autocatalytic Conversion of 1-Methyl-r-1,t-2-cyclopropanediamine into 4-Aminobutan-2-one in Aqueous Buffer Solutions of pH 6.5-10. Azomethine Ylides as Intermediates in the Catalysis by Alkyl Ketones

Article abstract of DOI:10.1021/jo952172y

The pK_a values of 1,2-cyclopropanediammonium dibromides (1-5*2HBr) were determined by potentiometric titration with KOH.Corrections for overlapping ionizations and for the varying ionic strength were employed in the calculation of the thermodynamic acidities.The equilibrium constant K_E = 2.9 of the two monoprotonated forms of trans-3b*H⁺ was derived from the pK_a values and the pH dependence of proton spectra which were interpreted with the help of nuclear Overhauser experiments.The autocatalytic conversion of trans-3a*2HBr into 4-aminobutan-2-one (8) in various buffer solutions of pH = 6.5-10 was monitored by proton spectroscopy.This sequence of reactions is subject to a general acid/base catalysis and also catalyzed by acetone and butanone.

Full text of DOI:10.1021/jo952172y

4920
J . Org. Chem. 1996, 61, 4920-4928
Cyclop r op a n ed ia m in es. 6.1 Au toca ta lytic Con ver sion of
1-Meth yl-r -1,t-2-cyclop r op a n ed ia m in e in to 4-Am in obu ta n -2-on e in
Aqu eou s Bu ffer Solu tion s of p H 6.5-10. Azom eth in e Ylid es a s
In ter m ed ia tes in th e Ca ta lysis by Alk yl Keton es†
Wolfgang von der Saal*,2 and Helmut Quast*
Institut fu¨r Organische Chemie der Universita¨t Wu¨rzburg, Am Hubland, D-97074 Wu¨rzburg
Received December 6, 1995X
The pKa values of 1,2-cyclopropanediammonium dibromides (1-5‚2HBr) were determined by
potentiometric titration with KOH. Corrections for overlapping ionizations and for the varying
ionic strength were employed in the calculation of the thermodynamic acidities. The equilibrium
constant KE ) 2.9 of the two monoprotonated forms of trans-3b‚H+ was derived from the pKa values
and the pH dependence of proton spectra which were interpreted with the help of nuclear
Overhauser experiments. The autocatalytic conversion of trans-3a ‚2HBr into 4-aminobutan-2-
one (8) in various buffer solutions of pH ) 6.5-10 was monitored by proton spectroscopy. This
sequence of reactions is subject to a general acid/base catalysis and also catalyzed by acetone and
butanone.
Ta ble 1. Th er m od yn a m ic Acid ity Con sta n ts of th e
1,2-Cyclop r op a n ed ia m m on iu m Dibr om id es 1‚2HBr to
5‚2HBr a n d of Som e Mod el Com p ou n d s in Wa ter a t 25 °C
In tr od u ction
Recently, we have shown that the 1,2-cyclopropanedi-
ammonium dibromides, 1‚2HBr to 5‚2HBr, are stable in
aqueous mineral acids. In phosphate buffer solutions of
pH ) 5.8 at 25 °C, however, 1-4 begin to decompose after
an initiation period of several hours, while 5 is stable
under these conditions.3 The products are unstable as
well, with the exception of 4-aminobutan-2-one (8), which
arises almost quantitatively from trans-3a ‚2HBr. In
order to provide a background for a detailed mechanistic
picture of these reactions, we now report the macroscopic
thermodynamic acidity constants of 1‚2HBr to 5‚2HBr
and the equilibrium constant of the two monoprotonated
forms of trans-3b‚H+. Subsequently, we detail the kinet-
ics of the conversion of trans-3a ‚2HBr into 8 and describe
the influence of various buffer solutions and ketones on
the rate constants of this autocatalytic reaction.
compound
pKa1
pKa2
cis-1a ‚2HBr
4.20
5.90
3.56
5.65
4.35
5.90
3.68
5.68
5.83
5.52
5.94
5.97
5.40
6.66
6.13
6.47
6.85
7.47
9.04
8.82
8.73
8.73
8.98
8.98
9.13
9.03
9.06
9.05
9.03
8.96
9.22
9.65
9.93
9.94
9.93
10.29
trans-1a ‚2HBr
cis-1b‚2HBr
trans-1b‚2HBr
cis-2a ‚2HBr
trans-2a ‚2HBr
cis-2b‚2HBr
trans-2b‚2HBr
trans-3a ‚2HBr
trans-3b‚2HBr
trans-4a ‚2HBr
trans-5a ‚2HBr
cis-1,2-cyclobutanediamine5
trans-1,2-cyclobutanediamine5
cis-1,2-cyclohexanediamine6
trans-1,2-cyclohexanediamine6
1,2-ethanediamine6
N,N′-dimethyl-1,2-ethanediamine6
were determined by potentiometric titration with 0.1 M
KOH of 0.05 M solutions of the diammonium dibromides
1‚2HBr to 5‚2HBr in carbon dioxide-free water. Special
measures were taken to avoid errors due to cis-trans
isomerization or decomposition during the titration. As
mentioned in the Introduction, such reactions are to be
expected in neutral solutions. Therefore, we determined
K1 by titrating the first deprotonation step in the range
of pH ) 3.0-6.5 and subsequently, using a freshly
prepared solution, K2 in the range of pH ) 8.5-10. For
the calculation of the thermodynamic constants from the
titration data, we used the iterative mathematical treat-
ment given by Albert and Serjeant.4 This method takes
into account overlapping ionizations and the ionic strength,
which varies during the titration. The thermodynamic
constants are listed in Table 1.
Meth od s a n d Resu lts
pKa Valu es of 1,2-Cyclopr opan ediam m on iu m Ion s.
The macroscopic acidity constants K1 and K2 at 25 °C
Equ ilibr iu m Con sta n t KT of th e Mon op r oton a ted
F or m s of 1-Meth yl-N,N′-d im eth yl-r -1,t-2-cyclop r o-
†Dedicated to Professor Lloyd M. J ackman on the occasion of his
70th birthday with great respect and affection.
(4) Albert, A.; Serjeant, E. P. The Determination of Ionization
Constants, 2nd ed.; Chapman and Hall: London, 1971.
X Abstract published in Advance ACS Abstracts, J une 1, 1996.
(1) Part
5
of this series: ref 3. The results are part of the
(5) The pKa values of 1,2-cyclobutanediamines were determined in
the same way as described here for 1,2-cyclopropanediamines: Werner,
M. Dissertation, University of Wu¨rzburg, 1985.
(6) Perrin, D. D. Dissociation Constants of Organic Bases in Aqueous
Solution; Butterworths, London, 1965; and Supplement, 1972.
Dissertation of W. von der Saal, University of Wu¨rzburg, 1983.
(2) Present address: Boehringer Mannheim GmbH, Department of
Chemistry, Sandhofer Str. 116, D-68305 Mannheim, Germany.
(3) von der Saal, W.; Quast, H. J . Org. Chem. 1995, 60, 4024.
S0022-3263(95)02172-4 CCC: $12.00 © 1996 American Chemical Society

Cyclopropanediamines
J . Org. Chem., Vol. 61, No. 15, 1996 4921
Sch em e 1
p a n ed ia m in e (tr a n s-3b). In aqueous solutions of a
diamine with different protonation sites, two tautomeric
monoprotonated species A+B and AB+ exist which are
connected by an equilibrium constant KT (eq 1).7 The
composite, i.e. macroscopic, acidity constants K1 and K2
are related to the microscopic constants, KA1, KA2, KB1
,
and KB2 by eqs 2-47 (Scheme 1).
For diamines with identical protonation sites, i.e. cis-
and trans-1, cis-2, trans-4a , and trans-5a , the microscopic
constants KA1 and KB1 are one-half of the macroscopic
constant K1, while KA2 and KB2 are twice as large as K2.
For diamines with nonequivalent amino groups, i.e.
trans-2 and trans-3, it is not obvious how to partition the
observed total acidities into those contributed by each of
the two different ammonium groups. The four micro-
scopic constants cannot be calculated from the macro-
scopic acidity data alone, because there are only three
equations (eqs 2-4). For the N,N′-dimethylcyclopro-
panediamine trans-3b, we obtained the missing informa-
tion, i.e. the size of KT (eq 1), by monitoring the change
of the chemical shifts of the N-methyl groups in proton
spectra recorded in the course of a titration.8
The N-methyl groups of trans-3b‚2HBr give rise to two
distinct singlets in proton spectra recorded from aqueous
solutions [2.86 and 2.93 ppm (D2O),9 2.80 and 2.86 ppm
(H2O)]. On the basis of the ABX spectrum of the ring
protons,9 we assigned the N-methyl signals with the help
of nuclear Overhauser experiments (see Experimental
Section), which demonstrated that the high-field signal
originates from the methylamino group at C1 of the
cyclopropane ring. The chemical shifts reflect the degree
of protonation of the methylamino groups and hence vary
with the pH of the solution: Both decrease when KOH
is added. Whereas the chemical shift of the C-methyl
group (at C1) changes linearly and with the same slope
F igu r e 1. Changes of the chemical shifts (ppm) of the methyl
groups at C1 (circles), at C1-N (squares), and at C2-N
(triangles) in 400-MHz proton spectra of trans-3b‚2HBr,
recorded from aqueous solutions at 25 °C, upon addition of
0.1 M KOH.
over the whole range, the chemical shift of one of the
N-methyl groups decreases faster than that of the other
until exactly 1 equiv of KOH has been added. Thereafter,
the chemical shifts behave inversely (Figure 1).
Obviously, the acidities of the two ammonium groups
A+ and B+ are different. The methylammonium group
at C1 and referred to as A+ in the following is less acidic
because its chemical shift decreases more slowly on
addition of KOH. For the evaluation of the tautomeric
constant KT, we considered this chemical shift first. The
various forms of a diamine, differing in the degree and,
in case of the monoprotonated species, the site of proto-
nation, equilibrate rapidly. Therefore, the observed
chemical shifts are the weighted averages of the chemical
shifts of the equilibrating species. On this basis, Hine
et al.8 derived eq 5.10 The sum of the fractions of the
+
+
monoprotonated species, i.e. fA + fAB , and the fraction
B
+B+
of the diprotonated form (fA ) can be calculated from
the macroscopic constants and the pH value.11 A plot of
+B
+
the left side of eq 5 vs fA + fAB yields a straight line
through the origin with slopeA (eq 6):
(7) Equations 1-4 are from ref 4.
KT ) [A+B]/[AB+]
(1)
(2)
(3)
(4)
fA
+
B
slopeA )
(∆δAA - ∆δAAB+) + ∆δA
(6)
+
B
+
AB
fA + fAB
+
+
B
K1 ) KA1 + KB1
Second, eq 7 was derived in the same way by considering
the chemical shift of the more acidic methylammonium
1/K2 ) 1/KB2 + 1/KA2
K1K2 ) KA1KB1 + KA2KB2
(8) Hine, J .; Via, F. A.; J ensen, J . H. J . Org. Chem. 1971, 36, 2926.
(9) von der Saal, W.; Reinhardt, R.; Seidenspinner, H.-M.; Quast,
H. Liebigs Ann. Chem. 1994, 569.
(10) The chemical shifts ∆δ are measured downfield from the signal
of the free base AB:
(11) The fractions are calculated according to the following:4
K1K2
fAB
)
[H+]2 + K1[H+] + K1K2
∆δAobs - fA +∆δA
)
A+B+
+
K1[H+]
[H+]2 + K1[H+] + K1K2
B
fA
(fA + fAB+) )
+
B
+
B
(fA + fAB
)
(∆δAA - ∆δAAB+) + ∆δA
(5)
+
+
+
+
AB
B
B
[
]
(fA + fAB
)
+
+
B
∆δAobs denotes the observed chemical shift, ∆δAA
the shift of A+B+,
[H+]2
+
+
B
∆δAA the (unknown) shift of A+B, and ∆δAAB (also unknown) that of
fA
+
)
+
B
+
+
B
[H+]2 + K1[H+] + K1K2
AB+. fA , fA B, and fAB denote the corresponding fractions.
+
+
+B+

4922 J . Org. Chem., Vol. 61, No. 15, 1996
von der Saal and Quast
A
A
B
F igu r e 3. Chemical shifts (ppm) of the N-methyl groups in
the proton spectra recorded from aqueous solutions of trans-
N,N′-dimethylcyclopropanediammonium dibromide (trans-1b‚
2HBr) at 25 °C after addition of 0.1 M KOH.
+
+
+ +
B
F igu r e 2. ∆δobs - fA ∆δA+B+ (circles) and ∆δobs - fA
B
∆δBA+B+ (squares) for 1-methyl-N,N′-dimethyl-r-1,t-2-cyclo-
+
+
propanediamine (trans-3b) vs the sum fA + fAB of the
B
fractions of the monoprotonated species.
In order to obtain a more realistic value for KT, we
derived the effect of remote protonation from a model
compound. Hine et al.8 reported that, in a 60-MHz proton
spectrum, the methoxy signal of 2-methoxyethanamine
shifts by 2 Hz upon protonation of the amino group. We
employed this value (13 Hz at 400 MHz) as an estimate
for the effect of protonation of the remote amino group,
because the observed protons are separated by five bonds
from the protonation site in both, 2-methoxyethanamine
and trans-3b. In addition, we assumed the protonation
shifts brought about by the remote amino groups to be
the same for A and B:
group B+ at C2:
fAB
+
slopeB )
(∆δBAB+ - ∆δB B) + ∆δAB
(7)
+
A
+
B
fA + fAB
+
+
B
Plots in this manner of the experimental data yield slopeA
) 137 Hz and slopeB ) 61 Hz (Figure 2).
The tautomeric constant KT (eq 1) is the ratio of the
fractions of the monoprotonated species, and hence eq
10 results.12
fA B (slopeA - ∆δAAB+)(∆δB + - ∆δB
)
)
+
+
A B
AB
KT )
)
(10)
(slopeB - ∆δAB B)(∆δAA B - ∆δAAB
fAB
+
+
+
+
∆δAAB+ ) ∆δB ) 13 Hz
(11)
+
A B
In order to be able to calculate KT from eq 10, we needed
Second, we considered that, most probably, ∆δAA is
estimates for the four unknown chemical shifts:
+
B
different from ∆δBAB+. To account for this difference, we
assumed that the shift resulting from protonation of the
observed methylamino group is the same whether (right
sides of eq 12 and 13) or not (left sides of eq 12 and 13)
the remote amino group is already protonated:
∆δAA B, ∆δAAB+, ∆δB B, and ∆δABB
.
+
+
A
+
The chemical shift of an N-methyl group is mainly
determined by the partial charge at the nitrogen atom
to which it is attached and much less by the charge of a
remote ammonium group. In other words, the change
in chemical shift brought about by protonation of the
amino group bearing the observed methyl group
∆δAA B ) ∆δAA B+ - ∆δA
(12)
(13)
+
+
+
AB
(∆δAA or ∆δABB+) is much larger than that caused by
+
B
protonation of the remote amino group (∆δAAB or
+
∆δBAB+ ) ∆δB + - ∆δB
+
A B
+
A B
∆δBA ). In order to derive a rough estimate of KT, we
+
B
neglected the effects of protonation of the remote amino
The validity of this approximation may be questioned.
One may speculate, for instance, that on protonation of
an amino group the cyclopropane ring may undergo
structural changes that influence the chemical shift of
the N-methyl groups. Protonation of the observed amino
group will thus give rise to different effects if the remote
amino group is protonated or not. We addressed this
question with the help of the symmetrical cyclopro-
panediammonium dibromide trans-1b‚2HBr as a model.
A plot of the chemical shift of the N-methyl groups vs
the amount of added KOH yielded a straight line (Figure
3). This proved that the assumptions made in eqs 12 and
13 are valid. Otherwise one would observe a point of
inflection in the line of Figure 3 after the addition of 1
equiv of KOH.
groups, i.e. ∆δAAB ) ∆δAB ) 0, and equated the effects
+
+
B
of protonation of the amino groups on the chemical shift
of their own methyl group, i.e. ∆δAA ) ∆δABB
. With
+B
+
these approximations, eq 10 yielded KT as ratio of the
slopes in Figure 2: KT ) 137/61 ) 2.3.
(12) Expressions for the unknown fractions of both monoprotonated
forms can be derived by rewriting eqs 6 and 7:
(slopeA - ∆δAAB+)(f
+ fAB
B
)
+
+
A
fA
)
)
(8)
(9)
+
B
(∆δAA - ∆δAAB
)
+
+
B
(slopeB - ∆δAB B)(fA + fAB
)
+
+
B
+
fAB
+
(∆δBAB+ - ∆δB
)
B
+
A

Cyclopropanediamines
J . Org. Chem., Vol. 61, No. 15, 1996 4923
Sch em e 2
Using the experimental values for ∆δAA
(199 Hz)
+
+
B
and ∆δBA
(223 Hz) and the estimated value for ∆δAAB
+
B
+
+
(13 Hz), we calculated KT ) 2.9 (eq 1413). If the effect of
protonation of the remote amino group was neglected,
i.e. ∆δAAB+, then KT ) 2.5.
F igu r e 4. Conversion of 1-methyl-r-1,t-2-cyclopropanediam-
monium dibromide (trans-3a ‚2HBr, circles, 0.05 M solution in
0.5 M phosphate buffer of pH ) 7.4 at 35 °C, run 8) into
4-aminobutan-2-one (8, squares). The curves were drawn by
a nonlinear fit14 of the data to eq 17.
With the knowledge of the tautomeric constant KT, a
detailed picture of the equilibria can be derived including
the calculation of the microscopic acidity constants
(Scheme 2).
Kin etic Measu r em en ts of th e Con ver sion of tr a n s-
3a .2HBr in Aqu eou s Bu ffer Solu tion s. The conver-
sion of trans-3a ‚2HBr into 8 in various buffer solutions
at 35 °C was monitored by proton spectroscopy with the
help of the signals of the methyl groups. The results are
listed in Table 2.
The reaction started after an initiation period, which
lasted approximately 1 h at pH ) 5.8 and 25 °C,3 but
only a few minutes at pH ) 8 and 35 °C. Once a small
amount of 8 had arisen, the rate increased considerably.
These observations were indicative of a mechanism
involving a slow, “spontaneous” conversion of trans-3a
into 8 with a rate constant k1 followed by a rapid
autocatalytic conversion into 8 with a rate constant k2.15
This kinetic scheme yielded eq 15 and its integrated form,
eq 16, where [ABt] denotes the total amount of the
cyclopropanediamine present, i.e. the free base [AB], the
monoprotonated forms [A+B] and [AB+], and the cyclo-
propanediammonium ion [A+B+], and [C] the concentra-
tion of 8 which is the catalyst.
If the catalyst is absent at the onset ([C]0 ) 0), eq 16
is reduced to eq 17. A nonlinear fit14 of the data in Figure
(13) The combination of eqs 10-13 yields eq 14:
(slopeA - ∆δAAB+) (∆δB + - 2∆δA
)
)
+
A
+
B
AB
KT
)
(14)
(slopeB - ∆δAAB+) (∆δA + - 2∆δA
+
A
+
B
AB
(14) The program used was SigmaPlot from J andel Scientific,
D-40699 Erkrath, Germany.
(15) Moore, J . W.; Pearson, R. G. Kinetics and Mechanism, J . Wiley
and Sons: New York, 1981; p 26. The kinetic scheme and the equations
are as follows:
F igu r e 5. (a) Variation of the second-order (“autocatalytic”)
rate constant k2 with the pH value in various buffer solu-
tions: open symbols, amine buffers; filled circle, phosphate
buffers; filled square, borate buffers. The second-order regres-
sion curve, calculated from the data of the amine buffers, is
shown. The data of runs 2, 19, 21, and 23 in Table 2 have been
omitted. (b) Variation of the composition of a 0.05 M solution
of 1-methyl-r-1,t-2-cyclopropanediammonium dibromide (trans-
3a ‚2HBr) between pH ) 6 and 10: dashed line, concentration
of the dication trans-3a ‚2H+; solid line, sum of the concentra-
tions of the two monoprotonated forms ([A+B] + [AB+]); dotted
line, free base trans-3a .17
k1
98 C
ABt
k2
98 2C
ABt + C
-d[ABt]
dt
) k1[ABt] + k2[ABt][C]
(15)
(16)
[ABt]0{k1 + k2([ABt]0 + [C]0)}
k2[ABt]0 + (k1 + k2[C]0)e{k +k ([AB ] +[C] )}t
[ABt]t )
1
2
t
0
0
[ABt]0(k1 + k2[ABt]0)
[ABt]t )
t 0
4 (run 8) to eq 17 yielded the rate constants k1 ) (3.1 (
(17)
k2[ABt]0 + k1e(k +k [AB ] )t
1
2
0.2) ×10-5 s-1 and k2 ) (2.09 ( 0.06) × 10-2 L mol-1 s-1 16
.

4924 J . Org. Chem., Vol. 61, No. 15, 1996
von der Saal and Quast
Ta ble 2. Exp er im en ta l Con d ition s a n d Kin etic Resu lts of th e Con ver sion of 1-Meth yl-r -1,t-2-cyclop r op a n ed ia m m on iu m
Dibr om id e (tr a n s-3a ‚2HBr , 0.05 M, 35 °C) in to 4-Am in obu ta n -2-on e (8) in Aqu eou s Bu ffer Solu tion sa
run
bufferb
[HB]c
[B-]d
[KCl]e
pH
f
g
h
105k1 (s-1
)
102k2 (L mol s-1
)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
MES
MES
0.20
0.40
0.20
0.15
0.10
0.075
0.05
0.05
0.025
0.40
0.20
0.30
0.20
0.20
0.20
0.15
0.10
0.17
0.17
0.025
0.25
0.20
0.20
0.10
0.30
0.60
0.30
0.35
0.40
0.425
0.45
0.45
0.475
0.10
0.30
0.20
0.30
0.30
0.30
0.35
0.40
0.33
0.33
0.475
0.25
0.30
0.30
0.40
1.20
0.90
1.20
0.30
0.20
0.15
0.10
0.10
0.05
0.70
1.20
0.90
1.00
1.20
1.20
1.20
1.30
1.17
1.17
0.75
1.25
1.20
1.20
1.10
6.4-6.5
6.4-6.5
6.8-6.9
6.8
7.0
7.2
7.4
7.4
7.6
2
1
0.5
3
3
2
2
2
2
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
1
1
1
1
1
B
B
B
A
A
A
A
B
A
C
C
C
C
C
C
C
C
A
B
A
B
A
B
B
35
30
11
50
50
60
60
50
50
10
10
10
10
10
10
10
10
10
30
30
20
30
30
50
7.8 ( 0.7
11.8 ( 1.2
48 ( 6
2.0 ( 0.1
3.0 ( 0.2
11.7 ( 1.0
1.55 ( 0.01
1.76 ( 0.03
1.99 ( 0.02
2.16 ( 0.02
2.09 ( 0.06
2.38 ( 0.04
96 ( 4
BIS-TRIS
phosphate
phosphate
phosphate
phosphate
phosphate
phosphate
TRIS
HEPES
TRIS
glyNH2
glygly
TAPS
TRIS
TRIS
gly
gly
NH4Cl
boric acid
boric acid
boric acid
boric acid
1.17 ( 0.03
1.23 ( 0.06
1.48 ( 0.04
1.29 ( 0.05
3.1 ( 0.2
4.2 ( 0.2
2.3 ( 0.7
5.7 ( 1.1
5.7 ( 0.3
6.1 ( 1.6
14 ( 2
3.1 ( 0.7
1.3 ( 0.2
4.0 ( 0.5
7 ( 4
7.4 ( 0.5
7.1 ( 1.0
2.8 ( 0.5
7.4 ( 0.7
13.6 ( 1.6
3.5 ( 0.1
7.5-7.8
7.8
76 ( 3
7.9-8.1
8.1-8.2
8.1-8.2
8.3-8.4
8.4-8.6
8.7-8.9
9.2-9.4
9.2-9.4
9.2-9.4
9.0-9.5
9.2-9.5
9.2-9.5
9.8-10
82 ( 7
39 ( 3
28.4 ( 1.4
60 ( 2
84 ( 2
49.4 ( 1.0
3.9 ( 1.1
4.2 ( 0.2
6 ( 3
5.9 ( 0.3
5.1 ( 0.2
4.0 ( 0.4
1.36 ( 0.03
1
2
a
b
The rate constants k1 and k2 (( asymptotic standard errors) were obtained by nonlinear fits14 of the data to eq 16.15 Concentration
0.5 M, ionic strength 1.5, abbreviations see Experimental Section. c Concentration of the acidic component of the buffer solution [M].
Concentration of the basic component of the buffer solution. e Concentration of KCl added for the adjustment of the ionic strength.
d
f Duration of the experiment (h). f Method for determination of the concentrations; A, sum of trans-3a and 8 ) 100%; B, comparison of the
signal heights of trans-3a and standard (sodium (trimethylsilyl)propanesulfonate); C, comparison of the signal heights of trans-3a with
g
the height of the same signal during the initial 2 min. Number of data used in the calculation.
Ca ta lysis by Bu ffer s. The rate constants k2 de-
pended on both pH and nature of the buffer. They
reached a maximum around pH 8 (Figure 5a). Amine
buffers were more effective than phosphate buffers. This
is quite evident when the value of k2 obtained from a
phosphate buffer at pH ) 6.8 is compared with that from
a BIS-TRIS buffer of the same pH (runs 3 and 4 in Table
2). The acceleration of the autocatalytic step by amine
buffers is even more clearly demonstrated by a compari-
son of the results obtained from a phosphate buffer and
a TRIS buffer of pH 7.6 (runs 9 and 10 in Table 2). In a
solution of the latter, the conversion was complete within
10 min, whereas several hours were required in a solution
of the former. The rate constants k2 also depended on
the concentration of the buffer. A 2-fold increase to a
concentration of 1 M accelerated the reaction consider-
ably (runs 1 and 2 in Table 2). These observations
indicate that the buffer (one or both components) par-
ticipates in the rate-determining step of the catalytic
cycle. The buffers act as general acid/base catalysts but
not as nucleophilic catalysts. This is inferred from the
observation that tertiary amine buffers (HEPES, run 11
in Table 2) gave rise to very similar rates as primary
amine buffers (TRIS, run 10).
In the experiments performed in phosphate buffer
solutions (runs 4-9 in Table 2), the rate constant k2
increases with increasing pH (6.8 f 7.6). Within this
pH range, the concentrations of both dication trans-3a ‚
2H+ and free base trans-3a are low. The main species
present are the two monocations trans-3a ‚H+, i.e. A+B
and AB+.17 Furthermore, the concentration of these
monocations is nearly constant in this pH range (Figure
5b). The related N,N ′-dimethyl monocations trans-
3b‚H+ have a tautomeric constant KT of 2.9. We expect
a similar behavior for trans-3a ‚H+ and hence that the
concentration of A+B is always larger than that of AB+.
The species involved in the rate-determining step
cannot be a combination of A+B+ and the general acid
- 18
H2PO4
,
because this is not compatible with the maxi-
mum of the rate constant near pH ) 8. The same is true
-
for a combination of the monocations and H2PO4
,
because the concentration of the monocations remains
constant between pH ) 6.8 and 7.6, and the concentra-
-
tion of H2PO4 declines, whereas k2 increases. A com-
-
bination of H2PO4 and the free base AB was found to
be consistent with the experimental results. The product
of the second-order rate constant k2 and the total
concentration of trans-3a can be replaced by the product
of a new, third-order rate constant k3, and the concentra-
tions of the general acid H2PO4 and the free base AB
(eq 18). Division of eq 18 by [ABt] affords an expression
(16) It may be assumed that the “spontaneous” ring opening does
not take place at all (k1 ) 0), but that a small amount of catalyst C is
present at the onset ([C]0 > 0). Thus, eq 16 collapses to
[ABt]0(k2[C]0 + k2[ABt]0)
[ABt]t )
t 0
-
k2[ABt]0 + k2[C]0e(k [C] +k [AB ] )t
2
0
2
which is identical with eq 17 if k1 in eq 17 is replaced by k2[C]0.
Consequently, a nonlinear fit of the data of Figure 4 to this equation
gives the same value for k2 as before and an estimate for [C]0 ) 0.0014
( 0.0006 M. Because the experimental design would not have allowed
detection of a catalyst at such a low concentration, we cannot rule out
that some 8 or another ketone may have been present already at the
beginning. Two recrystallizations of trans-3a ‚2HBr from methanol did
not affect the initiation phase, however. Therefore, eq 17 seems to
describe properly the conversion vs time curve, and we report in Table
2 rate constants k1 but not hypothetical initial concentrations of a
catalyst.
(17) The composition of the solutions at pH 6-10 (Figure 5b) was
calculated from the ionization constants K1 and K2 in the eq of footnote
11.
(18) The concentrations of the general acid H2PO4- and the general
base HPO42- were slightly different from those used for the preparation
of the buffer solutions, because addition of trans-3a ‚2HBr gave rise to
an increase of the pH value by 0.2-0.4. Therefore, we measured the
pH values and used the concentrations given for the measured pH
values by Green: Green, A. A. J . Am. Chem. Soc. 1933, 55, 2331.

Cyclopropanediamines
J . Org. Chem., Vol. 61, No. 15, 1996 4925
F igu r e 6. Variation of the rate constant k2 with the product
of the concentration of the general acid H2PO4 and the
fraction fAB of the free base trans-3a .
F igu r e 8. Concentration vs time curves for reactions of trans-
3a ‚2HBr (0.05 M solutions in 0.5 M phosphate buffer of pH )
6.8 at 35 °C) in the absence (filled circles) and in the presence
of butanone (open symbols; circles, 2 mM butanone; triangles,
3 mM; squares, 5 mM; diamonds, 8 mM).
-
Sch em e 3
F igu r e 7. Concentration vs time curves for reactions of trans-
3a ‚2HBr (0.05 M solutions in 0.5 M phosphate buffer of pH )
6.8 at 35 °C) in the absence (filled circles) and in the presence
of acetone (open symbols; circles, 2 mM acetone; triangles, 3
mM; squares, 4 mM; diamonds, 5 mM).
for a straight line through the origin with a slope k3 (eq
19).
k2[ABt] ) k3[H2PO4-][AB]
k2 ) k3[H2PO4-]fAB
(18)
(19)
A plot of experimentally determined rate constants k2 vs
the product of the concentration of the general acid
H2PO4- and the fraction of the free base (fAB) does indeed
yield a straight line through the origin (Figure 6) with a
slope k3 ) 6.5 L2 mol-2 s-1
.
Ca ta lysis by Aceton e a n d Bu ta n on e. We have
shown previously that the addition of minute amounts
of a solution, in which trans-3a ‚2HBr had completely
decomposed to afford 8, to a freshly prepared solution of
trans-3a ‚2HBr enhances the reaction rate dramatically3
(Scheme 3). The addition of ammonium chloride has no
effect. Even the use of ammonium chloride as component
of a buffer (run 20 in Table 2) does not yield a rate
constant which differs significantly from those obtained
with glycine (run 19) or borate (run 21) buffers. There-
fore, we assumed that the carbonyl moiety of 8 but not
its amino group participates in the rate-determining step.
Other ketones should hence behave similarly. Acetone
(Figure 7) and butanone (Figure 8) do indeed cause a
concentration-dependent increase of the rates. Therefore,
the previous kinetic scheme was extended to include the
reaction with ketones (eq 20).19
We defined a combined rate constant k4 (eq 21) which
depends on the chosen, constant ketone concentration.

4926 J . Org. Chem., Vol. 61, No. 15, 1996
von der Saal and Quast
panediamine may be stabilized by an intramolecular
hydrogen bond, which results in a decrease of pKa1 and
an increase of pKa2 as observed. N-Methyl groups reduce
pKa1 further by 0.6 for cis isomers and 0.2 to 0.3 for trans
isomers. This contrasts to the methyl group effect in the
case of the 1,2-ethanediammonium ion whose pKa1 is
increased by N-methyl groups by 0.6.6 N-Methyl groups
affect pKa2 much less and in a way which does not offer
an obvious consistent pattern.
The effect of a methyl group at C1 is revealed by a
comparison of the microscopic constants of trans-1b‚2HBr
(5.95, 8.43) with those of trans-3b‚2HBr (Scheme 2). The
methyl group decreases the acidity of the geminal and
increases that of the vicinal ammonium group.
Mech a n ism of th e Au toca ta lytic a n d th e Keton e-
Ca ta lyzed Rin g Op en in g. The nucleophilic reactivity
of amines toward carbonyl compounds parallels the
basicity of the amines.20 The free base trans-3a is hence
the most reactive species. The amino groups of both
monocations are less basic by 2-3 orders of magnitude.
Their reactivities are expected to be 50-100 times
smaller20 but their concentrations are 10-100 times
larger than that of the free base (Figure 5b). Therefore,
trans-3a and its monocations trans-3a ‚H+ may react with
similar rates to afford imines. Due to the small equilib-
rium constants for the formation of imines in aqueous
solution,21 direct observation of 6 and 9 by NMR spec-
troscopy was impossible. Similar systems of acid/base
equilibria as for trans-3a exist for the imines 6 and 9.
Again, the monoprotonated forms predominate but their
tautomeric equilibria are shifted far to the ammonium
ions because iminium ions are more acidic by 3 orders of
magnitude than the corresponding ammonium ions.22 For
the same reason, the dications are almost totally absent
and the proportions of the free bases are larger.
Most probably, it is the free base 6, which is attacked
by the general acid with concomitant ring opening to form
the azomethine ylide 7 in the rate-limiting step. The
azomethine ylide 7 is hydrolyzed to furnish the amino
ketone 8. As shown by the almost quantitative yield of
8 (95-98%), the analogous ring cleavage of the isomeric
imine 9, whose product 10 would be hydrolyzed to afford
aldehydes does not play a significant role. Only very
small amounts of aldehydes are actually formed in a
minor side reaction. These eventually yield the pyrroles
which are detected by the Ehrlich test.3
F igu r e 9. Dependence of the rate constant k4 on the concen-
tration of acetone (circles) and butanone (squares) according
to eq 21.
Ta ble 3. Ra te Con sta n ts k4 a n d k2 (( Asym p totic
Sta n d a r d Er r or ) Obta in ed by Non lin ea r F its of th e Da ta
of F igu r es 7 a n d 8 to th e In tegr a ted F or m of Eq 22
acetone
butanone
ketone
(mol L-1) run 104k4 (s-1) (L mol-1 s-1) run 104k4 (s-1) (L mol-1 s-1
102k2
102k2
)
0.002
0.003
0.004
0.005
0.008
25 1.18 ( 0.06 1.31 ( 0.06 29 0.62 ( 0.02 1.31 ( 0.03
26 1.29 ( 0.04 1.62 ( 0.05 30 0.78 ( 0.03 1.51 ( 0.05
27 1.70 ( 0.04 1.64 ( 0.04
28 2.36 ( 0.08 1.61 ( 0.08 31 1.24 ( 0.05 1.58 ( 0.06
32 1.76 ( 0.07 1.44 ( 0.07
Subsequently, we converted eq 20 into a rate equation
(eq 22) which is identical with eq 15. Thus, the inte-
grated form of eq 15, i.e. eq 16, could be employed, and
it yielded the rate constants k4 and k2 listed in Table 3.
k4 ) k1 + kketone[ketone]
-d[ABt]
(21)
(22)
) k4[ABt] + k2[ABt][C]
dt
The rate constant k2 remains essentially unchanged,
while k4 increases with increasing concentrations of the
ketones as expected. According to eq 21, plots of k4 vs
the ketone concentration afford straight lines with a
common intercept, which equals k1, and with slopes
corresponding to kketone (Figure 9). The slope kketone is 4.0
× 10-2 L mol-1 s-1 for acetone and 1.9 × 10-2 L mol-1 s-1
for butanone. The intercept is 2.4 × 10-5, a value which
is virtually identical with k1 obtained in the uncatalyzed
reactions. The observation that kketone of acetone is twice
as large as k2, the rate constant in the absence of acetone,
shows that acetone catalyzes more efficiently than 8.
Butanone is approximately as effective as 8.
If ring opening of 6 is the slow step, then 6 would have
to be formed from trans-3a and a ketone in a fast
equilibrium whose constant is smaller for an encumbered
(19) The kinetic scheme is
k1
98 C
ABt
k2
98 2C
Discu ssion
ABt + C
The relative configuration of the vicinal ammonium
groups at the cyclopropane ring strongly affects the pKa
values. cis-1,2-Cyclopropanediammonium ions are dis-
tinctly more acidic than the trans isomers (∆pKa1 ) pKa1-
(trans) - pKa1(cis) ) 1.6 to 1.7]. The (more flexible) 1,2-
cyclobutanediammonium ions (∆pKa1 ) 1.3)5 and 1,2-
cyclohexanediammonium ions (∆pKa1 ) 0.3)6 show the
same behavior but to a lesser extent. This difference
between the diastereomers can be traced back to the
proximity of the positive charges in the cis diammonium
ions. The monoprotonated form of a cis-1,2-cyclopro-
kketone
ABt + ketone
8 C + ketone
and eq 20 is
-d[ABt]
) k1[ABt] + k2[ABt][C] + kketone[ABt][ketone]
dt
(20) Hine, J . Acc. Chem. Res. 1978, 11, 1.
(21) The equilibrium constant for the imine formation from metha-
namine and acetone in water is 11.5 at 25 °C: Williams, A.; Bender,
M. L. J . Am. Chem. Soc. 1966, 88, 2508. On the basis of this constant,
the concentration of the imines would reach a maximum of 0.13 mM,
when the reaction had proceeded halfway.
(22) J encks, W. P. Acc. Chem. Res. 1976, 9, 425.

Cyclopropanediamines
J . Org. Chem., Vol. 61, No. 15, 1996 4927
ketone and hence also the rate of the reaction that is
catalyzed by this ketone. This is indeed borne out by the
experiments.
Exp er im en ta l Section
1H NMR Sp ectr a . Spectrometers used were WM 400 of
Bruker and EM 390 of Varian. Sodium 3-(trimethylsilyl)-
propanesulfonate was used as internal standard because its
shift does not change with the pH value of the solution.29
Ma ter ia ls. The 1,2-cyclopropanediammonium dibromides
were recrystallized from methanol/water (95:5) and dried in
vacuo over P2O5 for 1 d. The water employed in the experi-
ments was distilled twice in an all-quartz apparatus and
heated to 100 °C under N2 for 5 min prior to use. N2 used as
inert gas was bubbled through concentrated aqueous KOH and
subsequently through water. Standard buffer solutions were
purchased from Riedel-de Hae¨n, 0.100 M KOH from Merck.
Buffers were purchased from Sigma: 2-(morpholin-4-yl)-
ethanesulfonic acid (MES), 2-[N,N-bis(2-hydroxyethyl)amino]-
2-(hydroxymethyl)propane-1,3-diol (BIS-TRIS), 2-amino-2-
(hydroxymethyl)propane-1,3-diol (TRIS), 2-[4-(2-hydroxyethyl)-
piperazin-1-yl]ethanesulfonic acid (HEPES), 3-[2-hydroxy-1,2-
bis(hydroxymethyl)ethylamino]propane-1-sulfonic acid (TAPS),
glycine (gly), glycylglycine (glygly), and glycinamide hydro-
chloride (glyNH2).
The kinetic measurements do not unequivocally prove,
of course, that the rate-determining step is truly the
general acid-catalyzed ring opening of 6. (1) The reaction
of a general base with one or both monocations 6(A+B)
and 6(AB+) cannot be distinguished on the basis of
kinetics from the reaction of a general acid with the free
base 6 advocated above.23 (2) In the reaction of a weakly
basic amine with a ketone, the dehydration of the
intermediate adduct is catalyzed by acids and may be
rate-limiting.24 In this case, the rate law for the ketone-
catalyzed reaction would be eq 23.25 One would expect
encumbered ketones to react more slowly, as is indeed
observed. Therefore, we cannot exclude that formation
of the imine 6 is rate-limiting.
Con clu d in g Rem a r k s
Solutions (0.5 M) of MES, HEPES, TAPS, glycine, glycina-
mide hydrochloride, glycylglycine, NH4Cl, and boric acid were
prepared by dissolving the acidic forms and KCl in water and
addition of the appropriate amounts of 1 M NaOH and water.
Solutions of TRIS and BIS-TRIS buffers (0.5 M) were prepared
from the basic forms, KCl, and 1 M HCl. Phosphate buffer
(0.5 M) was prepared from KH2PO4, Na2HPO4, and KCl (Table
2).
p H Va lu es were measured with a glass electrode EM 125
containing an Ag/AgCl/KCl (3M) standard electrode and a
digital pH meter, both of Deutsche Metrohm GmbH, D-70794
Filderstadt, Germany. Prior to each measurement in the
range of pH 3-7, the electrode was calibrated at 25 °C with a
phosphate (pH ) 6.865) and a potassium phthalate buffer (pH
) 4.008). For the range between pH 7 and 10 the phosphate
buffer (pH ) 6.865) and a sodium tetraborate buffer (pH )
9.180) were employed.
Deter m in a tion of p Ka Va lu es. Glass electrode, standard
buffer solutions, and the apparatus containing 50.0 mL water
were kept for 5 h at 25.0 ( 0.1 °C. The apparatus consisted
of a 100-mL three-necked flask, magnetic stirrer, the glass
electrode, and a calibrated 5-mL glass buret connected to a
stainless steel syringe, and the whole was flushed with N2 for
0.5 h. A positive pressure of N2 was maintained throughout.
The buret was filled with 0.100 M KOH, which was kept in
the original plastic bottle as delivered.
Deter m in a tion of Ka for th e F ir st Dep r oton a tion Step .
A sample of the 1,2-cyclopropanediammonium dibromide
(0.250 mmol) was dissolved in water (50.0 mL). The syringe
was flushed with 0.100 M KOH from the buret, and the tip of
the syringe was placed a few millimeters above the surface of
the magnetically stirred solution. KOH (0.100 M, 0.50 mL)
was added to the solution, stirring was interrupted, and the
pH value was recorded after attaining a constant value, which
took 0.2-1 min. Stirring was started again, and a total of 10
portions (0.10-0.15 mL) of 0.100 M KOH was added. After
each addition, the pH value was recorded as described.
Det er m in a t ion of Ka for t h e Secon d Dep r ot on a t ion
Step . KOH (0.100 M, 3.00 mL) was added to water (50.0 mL),
followed by a sample of the 1,2-cyclopropanediammonium
dibromide (0.250 mmol). The solution was titrated as before.
Each compound was titrated twice.
The pKa values of the 1,2-cyclopropanediammonium
dibromides 1-5 and the equilibrium constant of the two
monocations of trans-3b help to interpret reactions in
aqueous solution.3 Furthermore, the data may serve to
improve computational methods for the prediction of pKa
values of small diamines. Such methods have recently
regained much interest in chemistry and biology.26 Up
to now, the mathematical models had to include assump-
tions about the principal conformations, because pKa
values of only flexible molecules were available.27
The conversion of trans-3a ‚2HBr into 8 in aqueous
buffer solutions of pH 6.5-10 is an autocatalytic process.
In addition, it is catalyzed by acetone or butanone and
is subject to general acid/base catalysis, particularly by
amine buffers. The reactive species is an imine 6, which
opens the ring to an intermediate 7 possessing both an
azomethine ylide28 and an imine moiety.
The proposed mechanism may also operate in the
decomposition of other 1,2-cyclopropanediamines which
commences in aqueous phosphate buffers only after a
similar initiation period. Their products are too unstable
to be observed, however. Aldehydes do not catalyze the
conversion of trans-3a into 8 but do furnish pyrroles.
Probably, this reaction also involves azomethineylides of
type 7 which, instead of being hydrolyzed, close a five-
membered ring. This formation of pyrroles from 1,2-
cyclopropanediamines will be dealt with in a forthcoming
paper.
(23) Lowry, T. H.; Richardson, K. S. Mechanisms and Theory in
Organic Chemistry, 2nd ed.; Harper and Row: New York 1981, p 608.
The combination of the acidity constants of the monocations trans-
3a ‚H+ and H2PO4- with eq 18 results in a rate equation for the reaction
of the general base HPO4 with the monocations trans-3a ‚H+.
2-
(24) J encks, W. P. Catalysis in Chemistry and Enzymology;
McGraw-Hill: New York, 1969; p 490.
(25)
d[ABt]
-
) kimine[ABt][HB][ketone]
(23)
dt
Ca lcu la tion of th e p Ka Va lu es. A total of 10 data points
from the first deprotonation step and 10 from the second were
used for the calculation of the pKa values with the help of the
iterative programme by Albert and Serjeant.4 The standard
deviation of each pKa value was less than 0.04.
(26) For a review, see: Honig, B.; Nicholls, A. Science 1995, 268,
1144.
(27) Potter, M. J .; Gilson, M. K.; McCammon, J . A. J . Am. Chem.
Soc. 1994, 116, 10298.
(28) Reviews on azomethine ylides: Tsuge, O.; Kanemasa, S. Adv.
Heterocycl. Chem. 1989, 45, 231. Claus, P. K. In Houben-Weyl,
Methoden der Organischen Chemie, 4th ed.; Klamann, D., Hagemann,
H., Eds.; Thieme: Stuttgart, 1990; Vol. E14b, Part 1, p 74. Vedejs, E.
In Advances in Cycloaddition; Curran, D. P., Ed.; J ai Press: Green-
wich, CT, 1988; Vol. 1, Kanemasa, S.; Tsuge, O. In Advances in
Cycloaddition; Curran, D. P., Ed.; J ai Press: Greenwich, CT, 1993;
Vol. 3. For more recent work, see for example: Ardill, H.; Grigg, R.;
Malone, J . F.; Sridharan, V.; Thomas, W. A. Tetrahedron 1994, 50,
5067 and references cited therein.
Equ ilibr iu m Con sta n t KT of th e Mon oca tion s tr a n s-
3b‚H+. Cyclopropanediammonium dibromide trans-3b‚2HBr
(152.0 mg, 0.550 mmole) was dissolved in water (10.0 mL) in
the apparatus used for the titrations. The pH value was
1
determined (3.989), and the H NMR spectrum was recorded
(29) De Marco, A. J . Magn. Reson. 1977, 26, 527.

4928 J . Org. Chem., Vol. 61, No. 15, 1996
von der Saal and Quast
at 25 °C from an aliquot of the solution (500 µL). KOH (0.100
35 °C. An NMR sample tube was charged with buffer solution
(500 µL), trans-3a .2HBr (6.21 mg, 0.025 mmol), and sodium
3-(trimethylsilyl)propanesulfonate (1.5 mg). Integration of the
signals was impossible because the concentration was too low
and the baseline was disturbed by the nearby water signal.
This signal also prevented the use of a pulsed NMR spectrom-
eter because it lead to detector overload at high amplification.
At low amplification, the signal-to-noise ratio was too bad
within the short periods of time (<1 min) required by the fast
rates of reaction. Therefore, the NMR spectra were recorded
on the CW spectrometer whose insert temperature (35 °C) was
determined before and after each experiment. The first
spectrum was taken 1.5 min after dissolving of trans-3a ‚2HBr
in the buffer solutions. Intervals of 10-30 s were maintained
between the scans. The heights of the methyl signals of trans-
3a , 8, and the standard were employed for the calculation of
the rate constants. Only the methyl signal of trans-3a was
used for the very fast reactions. The yield of 8, obtained from
trans-3a in phosphate buffer of pH ) 6.8 (95-98%), was
determined by cutting and weighting the methyl signals from
5 spectra (0.6 Hz/cm) recorded during the first 5 min and after
5 h at 35 °C.
The pH values were taken with a glass electrode from
samples that were identical with those of the NMR experi-
ments.
Ca ta lysis by Aceton e a n d Bu ta n on e. Acetone (145 µL)
was added to phosphate buffer solution (pH ) 6.8, 0.5 M, 50.0
mL). This solution (50 mM) was further diluted to 2, 3, 4,
and 5 mM by adding the appropriate amounts of buffer
solution. Solutions of butanone were prepared in the same
way.
M, 0.50 mL) was added from the buret followed by recording
1
of the pH and the H NMR spectrum at 25 °C. This procedure
was repeated seven times. The final pH value was 6.163. For
the second deprotonation step, the apparatus was charged with
water (10.0 mL) followed by 0.100 M KOH (4.50 mL) and trans-
3b‚2HBr (152.0 mg, 0.550 mmol). The procedure described
for the first deprotonation step was followed except that three
1-mL portions and, subsequently, three 0.5-mL portions of
0.100 M KOH were added. The initial pH value was 6.295,
the final 9.465. The results are shown in Figure 1.
The results shown in Figure 3 were obtained in the same
way from trans-1b‚2HBr.
NOE Exp er im en ts w ith tr a n s-3b‚2HBr . A solution (0.05
M) of trans-3b‚2HBr (9.66 mg) in D2O (0.7 mL) was kept at
25 °C in the insert of the 400-MHz NMR spectrometer. The
90° pulse angle was determined30 with the help of the N-
methyl singlets. T1 was estimated.31 2-H of the cyclopropane
ring showed the longest T1 (3.3 s). NOE difference spectra
were obtained by subtracting 400 spectra recorded with
selective saturation (decoupler attenuation 40 dB) of an
N-methyl group (90° pulse) and 400 spectra recorded with the
saturation frequency set outside the interesting area. Com-
plete recovery was achieved by a delay of 30 s after each pulse.
A spectrum, recorded with a sweep width of 2.3364 kHz and
16K data points in the time domain, pulse angle ) 90 °C, delay
) 30 s, and 8 pulses, afforded the phase correction used for
the NOE difference spectra.
Irradiation of the N-methyl group absorbing at 2.93 ppm
increased the signal of H2 (3.27 ppm) and H3 (1.44 ppm, cis to
the C-methyl group). Irradiation of the high-field N-methyl
signal (2.86 ppm) increased the signals of H2 and H4 (1.76 ppm,
trans to the C-methyl group). The signal of H2 showed the
largest enhancement in both experiments as expected, because
H2 is virtually equidistant to both N-methyl groups and has
the longest relaxation time.
Ack n ow led gm en t. We thank Professor Lloyd M.
J ackman, The Pennsylvania State University, for many
illuminative discussions. Financial support of this work
by the Deutsche Forschungsgemeinschaft and the Fonds
der Chemischen Industrie, Frankfurt am Main, is
gratefully acknowledged.
Kin etic Exp er im en ts. The buffer solutions were kept at
(30) Haupt, E. J . Magn. Reson. 1982, 49, 358.
(31) Martin, M. L.; Delpuech, J . J .; Martin, G. J . Practical NMR
Spectroscopy; Heyden and Sons: London, 1980; pp 246-275.
J O952172Y