Assessing the rates of ring-opening of aziridinium and azetidinium ions: A dramatic ring size effect

Article abstract of DOI:10.1021/ol200348k

Rates for the ring-opening of aziridinium and azetidinium ions by DMAP were measured. The four-membered ring appears to be ca. 17-000 times less reactive compared to the three-membered ring but is still highly relevant from a synthetic viewpoint. The elec

Full text of DOI:10.1021/ol200348k

ORGANIC
LETTERS
2011
Vol. 13, No. 7
1836–1839
Assessing the Rates of Ring-Opening of
Aziridinium and Azetidinium Ions: A
Dramatic Ring Size Effect
Nicolas De Rycke, Olivier David, and Franc-ois Couty*
ꢀ
Institut Lavoisier - Versailles, UMR 8180, Universite de Versailles St-Quentin-en-
Yvelines, 45 avenue des Etats-Unis, 78035 Versailles Cedex, France
couty@chimie.uvsq.fr
Received February 7, 2011
ABSTRACT
Rates for the ring-opening of aziridinium and azetidinium ions by DMAP were measured. The four-membered ring appears to be ca. 17 000 times
less reactive compared to the three-membered ring but is still highly relevant from a synthetic viewpoint. The electrophilicity of these strained
ammonium ions is measured for the first time.
The strain associated with small rings and its release
during a reaction are powerful driving forces for achieving
otherwise difficult transformations.1 In this field, aziridines
have become valuable substrates for the synthesis of nitro-
gen-containing molecules, but their intracyclic nitrogen
atom has to be activated in order to support the negative
charge created by the ring scission during nucleophilic
opening. Thus, N-acyl, N-carbamoyl, or N-tosyl aziridines,
in which the anionic charge on the nitrogen atom is stabi-
lized, have emerged as substrates of choice for this purpose.2
Another way to activate strained aza-heterocycles involved
in nucleophilic opening lies in the formation of an ammo-
nium ion, which can be achieved by simple protonation or
alkylation. However, the first option is precluded when
basic nucleophiles are considered, while the second one
gives rise to extremely electrophilic entities, which can only
be isolated and fully characterized in certain cases.3 There-
fore, N-alkyl aziridinium ions 1 are usually produced in situ,
most frequently through intramolecular substitution from
β-N,N-dialkylamino halides or sulfonates, and their subse-
quent nucleophilic opening is the key step in numerous
syntheses of biologically relevant molecules.4 Furthermore,
aziridinium ions are known to be the active electrophilic
species produced by nitrogen mustards. Due to the phar-
macological relevance of this important class of antitumoral
molecules, their reaction in physiological mediums and their
targeted nucleophilic sites involved in DNA alkylation have
been investigated in detail.5 Despite this, to the best of our
(4) (a) Oxenfor, S. J.; Moore, S. P.; Carbone, G.; Barker, G.; O’Brien,
P.; Shipton, M. R.; Gilday, J.; Campos, K. R. Tetrahedron: Asymmetry
2010, 21, 1563. (b) Metro, T. X.; Pardo, D. G.; Cossy, J. J. Org. Chem.
2007, 72, 6556. (c) Frizzle, M. J.; Caille, S.; Marshall, T. L.; McRae, K.;
Nadeau, K.; Guo, G.; Wu, S.; Martinelli, M. J.; Moniz, G. A. Org.
Process Res. Dev. 2007, 11, 215. (d) Couturier, C.; Blanchet, J.; Schlama,
T.; Zhu, J. Org. Lett. 2006, 8, 2183. (e) Revesz, L.; Blum, E.; Wicky, R.
Tetrahedron Lett. 2005, 46, 5577. (f) Gala, D.; Danahukar, V. H.;
Eckert, J. M.; Lucas, B. S.; Schumacher, D. P.; Zavialov, I. A.; Buholzer,
P.; Kubisch, P.; Mergelsberg, I.; Sherer, D. Z. Org. Process Res. Dev.
2004, 8, 754. (g) O’Brien, P.; Towers, T. D. J. Org. Chem. 2002, 67, 304.
(h) Chang, T. H.; Sharpless, K. B. Org. Lett. 2000, 2, 3555.
(1) Hesse, M. Ring Enlargements in Organic Chemistry; VCH:
Weinheim, New-York, Basel, Cambridge, 1991.
(2) For some recent reviews on aziridines, see: (a) Olsen, C. A.;
Franzyk, H.; Jaroszewsky, J. W. Eur. J. Org. Chem. 2007, 1717. (b) Bisol,
T. B.; Sa, M. M. Quim. Nova 2007, 30, 106. (c) Watson, I. D. G.; Yu, L. L.;
Yudin, A. K. Acc. Chem. Res. 2006, 60, 27. (d) Hu, X. E. Tetrahedron 2004,
60, 2701. (e) Cardillo, G.; Gentilucci, L.; Tolomelli, A. Aldrichimica Acta
2003, 36, 39. (e) Sweeney, J. B. Chem. Soc. Rev. 2002, 31, 247. (f) Mc Coull,
W.; Davis, F. A. Synthesis-Stuttgart 2000, 10, 1347.
(3) (a) Olah, G. A.; Szilagyi, J. J. Am. Chem. Soc. 1969, 91, 2949. (b)
Leonard, N. J.; Jann, K. J. Am. Chem. Soc. 1960, 82, 6418. (c) Leonard,
N. J.; Paukstelis, J. V. J. Am. Chem. Soc. 1964, 86, 821. (d) Kim, Y.; Ha,
H.-J.; Yun, S. Y.; Lee, W. K. Chem. Commun. 2008, 4363. (e) Piotrows-
ka, D. G.; Wroblewski, A. E. Tetrahedron 2009, 65, 4310.
(5) For recent theoretical data, see: (a) Mann, D. J. J. Phys. Chem. A
2010, 114, 4486. (b) Vasilescu, D.; Adrian-Scotto, M.; Fadiel, A.;
Hamza, A. J. Biomol. Struct. Dyn. 2010, 27, 465. (c) Shukla, P. K.;
Mishra, P. C.; Suhai, S. Chem. Phys. Lett. 2007, 449, 323. (d) Bhatta-
charyya, P. K.; Mehdi, C. Indian J. Chem., Sect. B 2005, 44, 1319.
r
10.1021/ol200348k
Published on Web 03/08/2011
2011 American Chemical Society

knowledge, direct measurement of the second-order rate
constant associated with the nucleophilic opening of an
aziridinium ion has not yet been reported.6 On the other
hand, our involvement in the chemistry of strained azetidi-
nium 2,7 i.e. the higher homologues of aziridinium ions, has
led us to note that these species are much less electrophilic
than 1, which can be at first explained by a lower ring strain
in the four-membered ring (Figure 1). Here again, no direct
rate measurement of the nucleophilic opening of azetidi-
nium ions has so far been published. Aiming at a better
understanding of the electrophilic character of these highly
relevant building blocks for the synthesis of nitrogen-
containing molecules, we report herein the kinetics measure-
ments of their opening with neutral nucleophiles. This
allows for the first time a direct comparison of the reactivity
of these strained heterocycles and an estimation of their
electrophilicity.
react regioselectively at the unsubstituted carbon atom,3d,7e
these salts are perfect candidates for comparative nucleo-
philic openings, since their only difference lies in the size of
the heterocycle. However, in order to compare the kinetics
for both substrates, we had to find the proper nucleophile
fulfilling the following requirements: (i) measurable rates for
both substrates (i.e., not too fast for the aziridinium
substrate); (ii) the nucleophile or product should display a
specific UV-vis absorption in order to be able to accurately
follow the reaction; and (iii) the nucleophile should react
irreversibly. We initially studied the nucleophilic opening
with anionic nucleophiles, such as thiophenoxides, phen-
oxides, and benzoates, but in the case of 5 reaction rates
were too fast and could not be accurately measured by
stopped-flow techniques. We therefore shifted to neutral
nucleophiles and found that 4-dimethylaminopyridine
(DMAP) reacted with both substrates at measurable rates
in acetonitrile9 to produce pyridinium trifluoromethanesul-
fonates 9 and 10 (Scheme 2).
Scheme 2. Nucleophilic Opening of 5 and 8 with DMAP
Figure 1. Aziridinium 1 and azetidinium 2 ions.
Aziridinium and azetidinium trifluoromethane sulfo-
nates 5 and 8 were prepared as depicted in Scheme 1. These
salts could be conveniently isolated after reaction of the
corresponding amines 4 and 7 with methyltrifluorometha-
nesulfonate and were stable enough to be stored without
appreciable degradation for weeks.
Furthermore, presence of a specific UV absorption band
in these compounds allowed accurate measurement of
their concentration. These pyridinium salts were isolated
in good yields, and no detectable byproducts were formed,
except in the case of aziridinium 5, for which the crude
reaction mixture showed a small amount (<5%) of a
regioisomer. Since pyridinium ion is a potential nucleo-
fuge, we heated both subtrates in refluxing acetonitrile for
a week, but without a significant change or detectable
production of the more substituted regioisomer. This is a
good indication for the irreversibility of the reaction,
because a similar experiment conducted with a chloride
atom as a leaving group led to complete isomerization
through thermodynamic control in the case of azetidinium
ions.10 The kinetics of nucleophilic opening with aziridi-
nium 5 was followed by UV spectrophotometry: equal
amounts of DMAP and 5 in acetonitrile were mixed in the
quartz cell, and the increase of absorbance of the produced
pyridinium 9 was followed against time at a given
temperature. Figure 2 shows an example of such data. After
integration of the kinetics rate law, a plot of (1/C0 - x) -
(1/C0), whereC0 denotes the initial concentration of DMAP
Scheme 1. Synthesis of Azetidinium and Aziridinium Triflates
In both cases, they were produced as single diastereoi-
somers, which is due to the preferred disposition of the
reacting lone pair, for steric reasons.8 Considering that alkyl
substituted aziridinium and azetidinium ions are known to
(6) Numerous kinetics data on the reaction rates of nitrogen mus-
tards, precursor of the electrophilic aziridinium with different nucleo-
philes, have been published. See inter alia: (a) Yang, H. Bull. Soc. Chim.
Belg. 1996, 105, 23. (b) Cullis, P. M.; Green, R. E.; Malone, M. E. J.
Chem. Soc., Perkin Trans. 2 1995, 7, 1503. (c) Kundu, G. C.; Shullek,
J. R.; Wilson, I. B. Pharmacol., Biochem. Behav. 1994, 49, 621. (d)
Hovinen, J.; Petterson-Fernholm, T.; Lahti, M.; Vilpo, J. Chem. Res.
Toxicol. 1998, 11, 1377.
(7) (a) Couty, F.; Kletskii, M. J. Mol. Struct. 2009, 908, 3368. (b)
Drouillat, B.; Couty, F.; David, O.; Evano, G.; Marrot, J. Synlett 2008,
9, 1345. (c) Couty, F.; David, O.; Drouillat, B. Tetrahedron Lett. 2007,
48, 9180. (d) Couty, F.; David, O.; Durrat, F. Tetrahedron Lett. 2007, 48,
1027. (e) Couty, F.; David, O.; Durrat, F.; Evano, G.; Lakhdar, S.;
Marrot, J.; Vargas-Sanchez, M. Eur. J. Org. Chem. 2006, 3, 479.
(8) (a) For azetidiniums: Couty, F.; Durrat, F.; Evano, G.; Marrot, J.
Eur. J. Org. Chem. 2006, 4214. (b) For aziridiniums: Vedejs, E.; Kendall,
J. T. J. Am. Chem. Soc. 1997, 119, 6941.
(9) Initial experiments were conducted in DCM, but reaction of the
solvent with DMAP led to erratic results; see: Rudine, A. B.; Walter,
M. G.; Wamser, C. C. J. Org. Chem. 2010, 75, 4292.
(10) Sivaprakasham, M.; Couty, F.; Evano, G.; Srinivas, B.; Sridhar,
R.; Rama Rao, K. Arkivoc 2007, (x), 71.
Org. Lett., Vol. 13, No. 7, 2011
1837

Figure 2. Example of monitoring absorbance A by UV-vis
spectrometry at λmax = 291 nm as a function of time of the
reaction between aziridinium salt 5 and DMAP at 15 °C in
CH3CN.
Figure 4. Example of determination of second-order rate con-
stants for the reaction of azetidinium salts 8 with 4-DMAP in
CD3CN (20 °C) by NMR. [DMAP]0 = [8]0 = 1.60 ꢀ 10-3
and x the concentration of 9, gave straight lines with the rate
constant as the slope (Figure 3). Each experiment was
conducted in duplicate or triplicate at different tempera-
tures (see Supporting Information). It should be noted that
the observed rate constants include the production of the
minor regioisomer, but this can be neglected due to the high
regioselectivity.
mol L-1. k = 1.4 ꢀ 10-4 L mol-1 s-1
.
3
3
3
was followed at different temperatures, and the Arrhenius
plot (Figure 5) gave values of ΔH# þ46.5 kJ mol-1 and
3
ΔS# þ165.4 J mol-1 K-1 for this reaction. The strong
3
3
positive value for ΔS# is in accordance with a SN2 reaction
involving a charge neutralization12 because the positive
charge of the produced pyridinium is delocalized. This value
is quite high and might be due to strong desolvatation in the
transition state. This strong positive value can account for a
dissymmetrical transition state in which the positive charge
of the intracyclic nitrogen is neutralized to a large extent.
Figure 3. Example of determination of second-order rate con-
stants for the reaction of aziridinium salts 5 with DMAP in
CH3CN (15 °C, λmax = 291 nm). [DMAP]0 = [5]0 = 3.60 ꢀ 10-5
mol L-1. k = 1.84 L mol-1 s-1
.
3
3
3
Being much slower, the kinetics of the ring-opening of
azetidinium ion 8 was followed by NMR at 293 K.
Integration of the decreasing signals at 6.58 ppm (H-2 in
DMAP) and growing signal at 7.84 ppm (same proton in
pyridinium) allowed accurate determination of the relative
concentrations versus time and thus determination of the
rate constant (Figure 4). These experiments afforded a
ratio of 1.65 ꢀ 104 for the rates of the nucleophilic opening
of 5 versus 8 at 293 K. These experiments were also
conducted with more nucleophilic11 4-pyrrolidinopyridine
(PPY) at 25 °C (see Supporting Information) and gave a
similar ratio of 1.75 ꢀ 104 for the rate constants.
Figure 5. Arrhenius plot for the reaction of 5 with DMAP.
DMAP is a highly studied nucleophile, and its N para-
meter, which measures the strength of a nucleophile, and s
parameter, which characterizes the sensitivity of the nu-
cleophile on variation of the electrophile, following the
three parameter eq 1,13 have been determined in acetoni-
trile when reacting with benzhydrilium cations.11
log k20 °C ¼ s(N þ E)
ð1Þ
In order to gain further insights on the activation para-
meters of this SN2 reaction, the nucleophilic opening of 5
(12) (a) Earley, J. E.; O’Rourke, C. E.; Clapp, L. B.; Edwards, J. O.
J. Am. Chem. Soc. 1958, 80, 3458. (b) Burns, J. T.; Leffek, K. T. Can.
J. Chem. 1969, 47, 3725.
(11) Brotzel, F.; Kempf, B.; Singer, T.; Zipse, H.; Mayr, H. Chem.;
Eur. J. 2007, 13, 336.
1838
Org. Lett., Vol. 13, No. 7, 2011

Assuming thatthe sparameter doesnotchangetoa large
extent when switching to different electrophiles, it is then
possible to estimate the E parameter values from eq 1 for
aziridinium 5 and azetidinium 9, which are respectively
-14.4 and -20.7. The E parameter varies in Mayr’s scale
from -20.5 (less electrophilic benzylidene malonate) to 6.0
(more electrophilic unstabilized benzylic cation). There-
fore, azetidinium ions rank low on the electrophilicity
scale, with an E parameter similar to the case for benzyli-
dene malonate, while aziridinium ions falls into the range
of the stabilized benzhydrilium cation.
this important difference in the rates is not a simple reflec-
tion of the lowered ring strain in azetidines versus aziridines,
which is estimated14 to be around 2.1 kcal mol-1: other
3
subtle parameters govern the boost in electrophilicity when
switching from azetidinium to aziridinium ions. In conclu-
sion, we have determined and compared for the first time the
direct rates for the irreversible nucleophilic opening of an
aziridinium and azetidinium ion: the latter was found to be
1.7ꢀ 104 times less reactive. From a practical viewpoint, the
lowered reactivity of azetidinium compared to aziridinium
ions makes them an easy to handle yet sufficiently reactive
species. Additionally, they remain underscored building
blocks for synthetizing nitrogen containing molecules. This
work provides a new tool for the determination of the
parameters influencing the electrophilicity of these species
which will be the subject of future report.
The huge difference in the rates of nucleophilic opening
between 5and 9corresponds to a ΔΔG# of 5.6 kcal mol-1 at
3
293 K, which is in agreement with recently calculated values
of 6.8 kcal mol-1 for the opening of aziridine and azetidine
with ammonia in the gas phase.14 It is clear, however, that
(13) General reviews: (a) Mayr, H.; Kempf, B.; Ofial, A. R. Acc.
Chem. Res. 2003, 36, 66. (b) Mayr, H.; Patz, M. Angew. Chem., Int. Ed.
Engl. 1994, 33, 938. (c) Mayr, H.; Patz, M.; Gotta, M. F.; Ofial, A. R.
Pure Appl. Chem. 1998, 70, 1993. (d) Mayr, H.; Ofial, A. R. In
Carbocation Chemistry; Olah, G. A., Prakash, G. K. S., Eds.; Wiley:
Hoboken, NJ, 2004; Chapter 13, pp 331-358. (e) Mayr, H.; Ofial, A. R.
Pure Appl. Chem. 2005, 77, 1807. (f) Lucius, R.; Loos, R.; Mayr, H.
Angew. Chem., Int. Ed. 2002, 41, 91. (g) Baidya, M.; Kobayashi, S.;
Brotzel, F.; Schmidhammer, U.; Riedle, E.; Mayr, H. Angew. Chem., Int.
Ed. 2007, 46, 6176. Nucleophilicities of arenes: (h) Mayr, H.; Bartl, J.;
Hagen, G. Angew. Chem., Int. Ed. Engl. 1992, 31, 1613. (i) Gotta, M. F.;
Mayr, H. J. Org. Chem. 1998, 63, 9769. (j) Herrlich, M.; Hampel, N.;
Mayr, H. Org. Lett. 2001, 3, 1629. (k) Herrlich, M.; Mayr, H.; Faust, R.
Org. Lett. 2001, 3, 1633.
Acknowledgment. N.D.R. thanks the University of
Versailles for funding. Pr. F. Terrier and Pr. R. Goumont
(University of Versailles) are acknowledged for fruitful
discussions.
Supporting Information Available. Synthesis and char-
acterization data for all new compounds. Data for the
kinetic measurements. This material is available free of
charge via the Internet at http://pubs.acs.org.
(14) Banks, H. D. J. Org. Chem. 2006, 71, 8089.
Org. Lett., Vol. 13, No. 7, 2011
1839