Mechanism of Selective Purine C-Nitration
A R T I C L E S
spectra as an abnormal intensity of the NMR signals; the signals
display either enhanced absorption or emission. The phase of
the net polarization observed, Γne (positive for enhanced
absorption signals, negative for emission signals), can be
predicted with Kaptein’s rules.47 When applied to the polariza-
tion of 15N nuclei, the equation takes the following form:48
Γne ) - µꢀai(gA - gB)
Here, µ is derived from how the radical pair is formed
(positive when formed from a triplet precursor or by the
encounter of free radicals, negative when formed from a singlet
precursor); ꢀ is derived from how the products are formed
(positive for reaction within the radical pair, negative for reaction
after separation of the radical pair); ai is the sign of the hyperfine
coupling constant for the observed nucleus i in radical A, and
(gA - gB) is the difference between the g values (obtained from
EPR data) of radicals A and B.
Figure 3. Conversion of N-nitro intermediate 11 to C2-nitro product 10
followed with H NMR at -10 °C. Closed circles, nitramine intermediate
11; open circles, product 10. kN, nitramine decrease; kP, product formation.
1
The excellent first-order kinetics confirm the expected uni-
molecular process. First-order rate coefficients, kN ) 1.5 × 10-3
s-1 (nitramine 11 decrease) and kP ) 1.9 × 10-3 s-1 (product
10 increase), were determined over about 4 half-lives with a
high R2 value and good reproducibility. The slight difference
in the values of kN and kP is explained by the occurrence of
side reactions (vide infra); a subsequent reaction would decrease
the magnitude of the signal for the product at the end of the
reaction and thus cause the extent of reaction at earlier times to
be overestimated.37
As both thermal38-40 and acid-catalyzed36,41 nitramine rear-
rangements of nitroanilines have been reported, we also allowed
the rearrangement to proceed in the presence of the base DIPEA,
and identical reaction rates were found (see Supporting Informa-
tion). This proved that the rearrangement was thermal and not
acid catalyzed. Moreover, if the rearrangement would be acid
catalyzed, the reaction rate would increase in the course of the
reaction, as TFA is generated upon product formation. Deviation
from a first-order correlation was not observed. To our
knowledge, this is the first identified example of a nitramine
rearrangement in a purine/pyrimidine system.
15N-CIDNP in Purine Nitration. With the unimolecularity
of the reaction now being established, we chose to study the
mechanism of the rearrangement with 15N NMR, which has been
a valuable tool in elucidating reaction pathways in nitration
reactions.42 Several mechanisms have been put forward for the
nitramine rearrangement, both heterolytic43 and homolytic
pathways.41 The observation of CIDNP effects in 15N NMR
spectra during the rearrangement of nitroaniline derivatives
offered convincing evidence that radicals were involved, thus
supporting a homolytic rearrangement mechanism.37,44 Chemi-
cally Induced Dynamic Nuclear Polarization refers to the
perturbation of the nuclear spins away from the expected
Boltzmann distribution.45,46 The effect is observed in NMR
In the case of the rearrangement of nitramine intermediate
11, the above equation can be reduced to the extent that the
sign of Γne discloses whether the rearrangement takes places in
an inter- or intramolecular fashion. In the NO2 radical, the
negative magnetogyric ratio of the 15N nucleus also causes aN
to be negative,49 while the analysis can be further simplified if
the reasonable assumption is made that the g value of the organic
purinyl radical50 is larger than the g value of NO2 (2.0000).49
The conclusions of Kaptein’s rules can now simply be related
to how the radicals are formed and how they react to form the
product. In a thermal radical rearrangement, the radical pair is
generated from a singlet precursor, so µ is negative. Thus, from
the phase of the polarized NMR signals, one can then conclude
whether the rearrangement occurs intra- or intermolecularly. If
the product is formed in an intramolecular fashion, it is the result
of a recombination reaction within the radical pair (ꢀ is positive)
and enhanced absorption is observed. If the product is formed
in an intermolecular fashion, it is the result of recombination
of the radicals after separation of the original pair (ꢀ is negative)
and emission is observed. Conclusions are less obvious if the
rearrangement has both an intra- and an intermolecular coun-
terpart. Since the observed NMR signal matches the sum of
the unpolarized and polarized (positive and/or negative) material
formed, even no net polarization at all can be the consequence.
When we followed the rearrangement of nitramine 11 at 0
°C with 15N NMR, indeed, CIDNP effects were observed (Figure
4). During several half-lives of the rearrangement, the doublet
of nitramine intermediate 11 at 339.6 ppm showed enhanced
absorption (see Supporting Information for real-time spectra).
In the early stage of the reaction at t ) 2 min, the singlet of
2-NO2 product 10 at 365.0 ppm showed an emission signal,
while during the remainder of the rearrangement, a reduced
(37) A similar discrepancy in the values of kN and kP was found by Ridd and
co-workers during the acid-catalyzed nitramine rearrangement of 2,6-
dichloro-N-nitroaniline and 2,6-dibromo-N-nitroaniline: Abu-Namous, A.
M. A.; Ridd, J. H.; Sandall, J. P. B. Can. J. Chem. 1986, 64, 1124-1129.
(38) Barnes, T. J.; Hickinbottom, W. J. J. Chem. Soc. 1961, 2616-2620.
(39) Naud, D. L.; Brower, K. R. J. Org. Chem. 1992, 57, 3303-3308.
(40) Naud, D. L. J. Chem. Soc., Perkin Trans. 2 1996, 1321-1324.
(41) (a) White, W. N.; Klink, J. R. J. Org. Chem. 1977, 42, 166. (b) White, W.
N.; White, H. S.; Fentiman, A. J. Org. Chem. 1976, 41, 3166-3170 and
earlier papers in this series.
(45) Lepley, A. R.; Closs, G. L. Chemically Induced Magnetic Polarization; J.
Wiley & Sons: New York, 1973.
(46) Muus, L. T.; Atkins, P. W.; McLaughlan, K. A.; Pedersen, J. B. Chemically
Induced Magnetic Polarization; D. Reidel Publishing Company: Dordrecht,
The Netherlands, 1977.
(47) Kaptein, R. J. Chem. Soc., Chem. Commun. 1971, 732-733.
(48) (a) Clemens, A. H.; Ridd, J. H.; Sandall, J. P. B. J. Chem. Soc., Perkin
Trans. 2 1984, 1659-1665. (b) The additional negative sign must be added
because of the negative magnetogyric ratio of the 15N nucleus. See: Porter,
N. A.; Dubay, G. R.; Green, J. G. J. Am. Chem. Soc. 1978, 100, 920-925.
(49) Morton, J. R.; Preston, K. F.; Strach, S. J. J. Phys. Chem. 1979, 83, 533-
536.
(50) Hellwege, K.-H.; Fischer, H. Landolt-Bo¨rnstein: Numerical Data and
Functional Relationships in Science and Technology; Springer-Verlag:
Berlin, 1977.
(42) Ridd, J. H. Chem. Soc. ReV. 1991, 20, 149-165.
(43) (a) Banthorpe, D. V.; Hughes, E. D.; Williams, D. L. H. J. Chem. Soc.
1964, 5349-5361. (b) Banthorpe, D. V.; Thomas, J. A.; Williams, D. L.
H. J. Chem. Soc. 1965, 6135-6140. (c) Brownstein, S.; Bunton, C. A.;
Hughes, E. D. J. Chem. Soc. 1958, 4354-4357.
(44) Ridd, J. H.; Sandall, J. P. B. J. Chem. Soc., Chem. Commun. 1982, 261-
262.
9
J. AM. CHEM. SOC. VOL. 127, NO. 16, 2005 5961