Nucleophilic reactivities of the anions of nucleobases and their subunits

Article abstract of DOI:10.1002/chem.201102411

The kinetics of the reactions of different heterocyclic anions derived from imidazoles, purines, pyrimidines, and related compounds with benzhydrylium ions and structurally related quinone methides have been studied in DMSO and water. The second-order rate constants (logk₂) correlated linearly with the electrophilicity parameters E of the electrophiles according to the correlation logk₂ = s_N(N+E) (H. Mayr, M. Patz, Angew. Chem. 1994, 106, 990-1010; Angew. Chem. Int. Ed. Engl. 1994, 33, 938-957) allowing us to determine the nucleophilicity parameters N and s_N for these anions. In DMSO, the reactivities of these heterocyclic anions vary by more than six orders of magnitude and are comparable to carbanions, amide and imide anions, or amines. The azole anions are generally four to five orders of magnitude more reactive than their conjugate acids. Copyright

Full text of DOI:10.1002/chem.201102411

FULL PAPER
DOI: 10.1002/chem.201102411
Nucleophilic Reactivities of the Anions of Nucleobases and Their Subunits
Martin Breugst, Francisco Corral Bautista, and Herbert Mayr*[a]
Dedicated to Professor Michael Hanack on the occasion of his 80th birthday
Abstract: The kinetics of the reactions
of different heterocyclic anions derived
from imidazoles, purines, pyrimidines,
and related compounds with benzhy-
drylium ions and structurally related
quinone methides have been studied in
DMSO and water. The second-order
rate constants (logk2) correlated linear-
ly with the electrophilicity parameters
E of the electrophiles according to the
correlation logk2 = s
(N+E) (H. Mayr,
and sN for these anions. In DMSO, the
reactivities of these heterocyclic anions
vary by more than six orders of magni-
tude and are comparable to carbanions,
amide and imide anions, or amines.
The azole anions are generally four to
five orders of magnitude more reactive
than their conjugate acids.
M. Patz, Angew. Chem. 1994, 106, 990–
1010; Angew. Chem. Int. Ed. Engl.
1994, 33, 938–957) allowing us to deter-
mine the nucleophilicity parameters N
Keywords: azole anions · kinetics ·
linear free energy relationships ·
pyrimidines · regioselectivity
Introduction
Imidazoles, pyrimidines, and purines as well as their deriva-
tives are omnipresent in chemistry, biology, and medicine
and are of tremendous importance for many syntheses.[1]
The imidazole moiety of histidine plays an important role in
the active center of several enzymes,[2] and Staab was among
the first to realize that imidazoles are very effective catalysts
in acylation reactions and ester hydrolyses.[3] While the neu-
tral imidazole is often used as a catalyst for the hydrolysis of
esters with good leaving groups (e.g., p-nitrophenyl acetate),
the imidazole anion is an effective catalyst for the hydrolysis
of esters with poor leaving groups (e.g., p-cresyl acetate).[1b,c]
Since imidazoles, purines, and pyrimidines are ambident
nucleophiles with several competing reaction centers, the
possibility of multiple reaction pathways complicates their
use in organic synthesis. A similar situation is found for
their anions, although attack along the dashed arrows in
Scheme 1, which is possible in the neutral compounds, can
be neglected for these anions due to the higher reactivity of
the negatively charged fragments.
Scheme 1. Conceivable reaction sites of the ambident heterocyclic anions
(solid arrows) and additional reaction sites of the neutral compounds
(dashed arrows).
resulting isomers depends on the electronic and steric effects
of the substituents.[1b,c] Thus, the methylation of 4-nitroimi-
dazole by dimethyl sulfate in aqueous NaOH occurs eight
times faster at N1 than at N3 (Scheme 2),[4] and selective N1
attack was observed in the reactions of 2-methyl-4-nitroimi-
dazole with alkyl halides or sulfates under alkaline condi-
tions indicating the higher nucleophilicity of N1 compared
to N3.[5]
Terrier and co-workers showed that under conditions of
kinetic control the imidazole anion is attacked by trinitro-
benzene at nitrogen. Subsequently, attack at C4 occurs to
yield the thermodynamically more stable product
(Scheme 3).[6]
The reactions of imidazoles with alkyl halides or dimethyl
sulfate under basic conditions, that is, the alkylations of the
imidazole anions, yield N-alkylated imidazoles. In the case
of unsymmetrically substituted imidazoles, the ratio of the
The regioselectivity is even more complicated for purine
derivatives. In extensive studies, the groups of Freccero and
[a] Dr. M. Breugst, M. Sc. F. Corral Bautista, Prof. Dr. H. Mayr
Department Chemie
Ludwig-Maximilians-Universitꢀt Mꢁnchen
Butenandtstrasse 5–13 (Haus F)
81377 Mꢁnchen (Germany)
Fax : (+49)89-2180-77717
E-mail: herbert.mayr@cup.uni-muenchen.de
Homepage: http://www.cup.uni-muenchen.de/oc/mayr
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201102411.
Scheme 2. Alkylation of the anion of 4-nitroimidazole with dimethyl sul-
fate in 90% water/10% ethanol at 258C.[4]
Chem. Eur. J. 2012, 18, 127 – 137
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
127

for medical applications, and several strategies involving
protective groups have been employed for the synthesis of
the desired alkylation products.[10] In reactions of uracil
anions, preferred N1-alkylation was observed with methoxy-
methyloxirane,[11] alkyl halides,[12] or lactones[13] in DMF;
also Michael reactions of uracil anions with acrolein resulted
in the predominant formation of N1-alkylated products.[14]
These selectivities are typically rationalized by the higher
acidity of the N1-H compared with the N3-H because of the
better delocalization of the negative charge in the N1
anion.[10] However, N1,N3-dialkylated compounds are typi-
cally formed as side products in these reactions.
Gambacorta and co-workers explained the different N/O
alkylation ratios in uracil derivatives with a qualitative hard-
ness scale (N1 < N3 < O4),[15] employing the HSAB princi-
ple[16] or the related Klopman–Salem concept of charge and
orbital controlled reactions.[17] However, we have recently
shown that the ambident reactivities of imide and amide
anions cannot generally be explained with these concepts.[18]
Free amide and imide anions are selectively attacked at ni-
trogen by benzhydrylium ions and quinone methides, and
the attack at the oxygen terminus does only occur when the
diffusion limit is reached or the nitrogen atom is blocked
(e.g., by silver ions).[18a,b]
The fact that only little quantitative data on the reactivi-
ties of these important classes of heterocycles is known in
the literature prompted us to study the nucleophilic reactivi-
ties of these compounds in detail. In earlier work, we have
shown that benzhydrylium ions and structurally related qui-
none methides (1) can be used as reference electrophiles
with tunable reactivity for characterizing a large variety of
nucleophiles.[19] The second-order rate constants at 208C of
the reactions of these nucleophiles have been described by
Equation (1),[20] where sN and N are nucleophile-specific pa-
rameters and E is an electrophile-specific parameter.
Scheme 3. Ambident reactivity of the imidazole anion towards trinitro-
benzene.[6]
Rokita examined the selectivities of the alkylation of purine
bases by the parent ortho-quinone methide (Scheme 4).[7]
Quantum-chemical calculations at B3LYP/6-311+GACTHUNTGRNEUNG(d,p)
level of theory predicted the following nucleophilic (i.e., ki-
netic) reactivity order for the reaction of adenine with the
quinone methide in water: N3 > N7 ꢀ N1 @ NH2, while a
different ordering is derived for the product stabilities
(NH2; after proton shift >N3 ꢀ N7 @ N1).[7c] The time-de-
pendent analysis of the adducts from deoxyadenosine with
the quinone methide showed a fast and reversible attack at
N1. However, a much slower reaction at the amino group
and subsequent proton transfer yield the thermodynamically
most stable reaction product (Scheme 4).[7d]
Some studies of purine anions showed that reactions at
the N1 and N3 positions do not occur. The anions of purine
nucleobases are exclusively attacked at N9 by epoxides,[8]
while dimethyl propargyl chloride attacks at N9 and N7 in
HMPT.[9]
The control of N1 versus N3 alkylation (see Scheme 1 for
numbering) in anionic uracil derivatives is very important
log k2 ¼ sN ðN þ EÞ
ð1Þ
We now report on the kinetics of the anions of several
imidazoles (2), purines (3), and pyrimidines (4) with the ref-
erence electrophiles 1 listed in Table 1 in DMSO and water,
and we show how these data can be used to include these
anions in our comprehensive nucleophilicity scale.[19f]
Results and Discussion
Reaction products: For elucidating the reaction pathways,
the quinone methides 1l and 1k were used as reaction part-
ners for the more reactive imidazole anions 2a–e, while the
benzhydrylium ion 1b was employed for less reactive nucle-
ophiles. Complete decolorization of the solutions, that is,
complete consumption of the electrophiles 1 were observed
in the product studies described in the following schemes,
when solutions of 1l,k or 1b were combined in DMSO with
1–5 equivalents of the potassium salts of 2–4 (exception: re-
action of 2e with 1l, Scheme 5). The fact that some reaction
Scheme 4. Alkylation of deoxyadenosine with the parent ortho-quinone
methide in water.[7d]
128
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 127 – 137

Nucleophilic Reactivities
FULL PAPER
Table 1. Reference electrophiles employed in this work and their wave-
lengths monitored in the kinetic experiments.
tography or recrystallization was carried out for these com-
pounds and the differentiation between different regioisom-
ers was based on 2D-NMR (COSY, HSQC, HMBC) experi-
ments.
Electrophile
E[a]
leval [nm]
R=N
R=NMe2
R=NACTHNUTGRNEUNG(CH2)4
G
ꢁ5.53
ꢁ7.02
ꢁ7.69
620
613
620
As depicted in Schemes 5 and 6, only one reaction prod-
uct was obtained in the reactions of symmetrical imidazole
anions. The reaction of the unsymmetrical anion of 4-meth-
ylimidazole (2c) with the quinone methide 1l resulted in a
2.5:1 mixture of two regioisomers in line with previous stud-
ies on the reactions with neutral azoles, where a similar
ratio for 4-methylimidazole 2c-H (2.5:1) was observed.[21] In
contrast to this observation, the anions of 2,4-dimethylimi-
dazole (2d), 4-nitroimidazole (2e), and 4-formylimidazole
(2g), that is, other unsymmetrically substituted anions, gave
rise to only one regioisomer (Schemes 5 and 6). The reac-
tion of 1b with the anion of 4-nitroimidazole (2e), which
was studied in [D6]DMSO by means of NMR-spectroscopy,
also gave only one reaction product (Scheme 6). From the
low stability of the reaction products discussed above, one
has to conclude that the products are obtained by thermody-
namic product control and do not necessarily reflect the ini-
tial kinetic distribution.
n=2
n=1
1d
1e
ꢁ8.22
ꢁ8.76
618
627
n=2
n=1
1 f
1g
ꢁ9.45
635
ꢁ10.04 630
R=OMe
R=NMe2
1h
1i
ꢁ12.18 422
ꢁ13.39 533
R=NO2
R=Me
R=OMe
R=NMe2
1j
1k
1l
ꢁ14.36 374
ꢁ15.83 371
ꢁ16.11 393
1m ꢁ17.29 486
[a]Electrophilicity parameters from ref. [19a,b,e].
These findings are in accordance with previous studies on
the reaction of the anion of 4-methylimidazole (2c) with tri-
nitrobenzene yielding a 4:1 ratio of N1- and N3-arylated 4-
methylimidazoles,[6b] or on the methylation of 4-nitroimida-
zole by dimethyl sulfate in aqueous NaOH, where a 9:1
ratio for N1 versus N3 was observed (Scheme 2).[4]
The reactions of the imidazole anions 2 with our reference
electrophiles (Schemes 5 and 6) yielded only the products of
nitrogen attack; C-alkylation of imidazole anions, as previ-
ously reported by Terrier for the reaction of 2a with trinitro-
benzene (Scheme 3),[6] was not observed in the reactions of
the imidazole anions 2a–e with the reference electrophiles
1. Kinetic control of N-attack can be explained by the
Scheme 5. Products of the reactions of the more reactive imidazole
anions 2a–e with the quinone methides 1k and 1l in DMSO at RT.
higher thermodynamic stability of 5aACTHNUTRGNEUNG(a–g) compared with 8
(Scheme 7), the nonaromatic initial product of C-alkylation,
and the lower intrinsic barrier for N-attack.[18c,22] The differ-
ent behavior of trinitrobenzene (Scheme 3) can be explained
by the higher electrofugality of trinitrobenzene compared
products were only obtained in moderate yields after aque-
ous or acidic work-up can be explained by non-optimized
work-up procedures or in the
case of the reaction of 2e with
1l by an equilibrium. While
the products formed from 2a–
d and the quinone methides 1k
and 1l are stable under the re-
action conditions, the reactions
of the potassium salts of the
less reactive anions with stabi-
lized benzhydrylium ions (e.g.,
1b) are reversible and the ini-
tially formed adducts 5–7 may
hydrolyse in aqueous solution
to yield the corresponding
benzhydrols as the final prod-
ucts. Therefore, no further pu-
rification by means of chroma-
Scheme 6. Products of the reactions of the less reactive imidazole anions 2e–g with the benzhydrylium ion 1b
in DMSO at RT.
Chem. Eur. J. 2012, 18, 127 – 137
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
129

H. Mayr et al.
the higher acidity of N1-H versus N3-H,[23] the anions of
uracil (4a) and thymine (4b) are attacked at the amide ni-
trogen N1. As N1 is blocked by a methyl group in 4c, selec-
tive alkylation takes place at the imide nitrogen N3. The dif-
ferent surrounding of the benzhydryl group in 7ab and 7bb
on one side and in 7cb on the other can also be derived
from the 13C NMR chemical shifts of the benzhydryl car-
bons, which resonate at d = 60.8 and 60.4 ppm for 7ab and
7bb and at d = 57.3 ppm for 7cb. These findings are in line
with the previously published products of the reactions of
uracil anions with a variety of electrophiles.[11–14] However,
we cannot exclude that the anions of uracil and thymine ini-
tially give mixtures of regioisomers with 1b, which subse-
quently isomerize to the isolated products 7ab and 7bb.[23b–d]
Scheme 7. C- versus N-alkylation of the imidazole anion 2a with benzhy-
drylium ions in DMSO.
with those of the studied benzhydrylium ions, which makes
N-alkylation in Scheme 3 more reversible and gives rise to
C-alkylation under conditions of thermodynamic control.
The reactions of the symmetrical anions of benzimidazole
(3a) and benzotriazole (3b) with the benzhydrylium ion 1b
Kinetic investigations: The reactions of the heterocyclic
anions 2–4 with benzhydrylium ions 1a–g and structurally
related quinone methides 1h–m were performed in DMSO
(Table 2) and water (Table 3) at 208C and were monitored
by UV/Vis spectroscopy at or close to the absorption
maxima of the electrophiles (Table 1). To simplify the evalu-
ation of the kinetic experiments, the nucleophiles were gen-
erally used in large excess over the electrophiles. Therefore,
the concentrations of 2–4 remained almost constant
throughout the reactions, and pseudo-first-order kinetics
were obtained in all runs. The first-order rate constants kobs
were then derived by least-squares fitting of the time-depen-
dent absorbances At of the electrophiles to the exponential
gave rise to
a single reaction product, respectively
(Scheme 8). While the reaction of the unsymmetrical purine
anion 3c resulted in a 1.2:1 mixture of the N7- and the N9-
alkylated purines, the anions of theophylline (3d) and ade-
nine (3e), which are also unsymmetrical, reacted regioselec-
tively with 1b to give uniform products 6bd and 6eb, re-
spectively.
N-Alkylated pyrimidines (7) are formed exclusively by
the reactions of the anions 4a–c with the benzhydrylium ion
1b as revealed by 2D HMBC spectroscopy (Scheme 9). In
analogy to the behavior of other imide and amide an-
ions,[18a,b] O-alkylation has not been observed in the reac-
tions of 4a–c with the benzhydrylium ion 1b. In line with
function At =A0exp
A
stants were obtained as the slopes of the linear plots of kobs
versus the concentrations of the nucleophile (Figure 1).
For the investigations in DMSO solution, the potassium
salts of the nucleophiles 2–4 were used. It has been shown
that kobs values obtained in the
presence and in the absence of
crown ether are on the same
kobs versus [2–4] plots (see Sup-
porting Information) indicating
that the reactivities of the free
anions were determined.
Due to their high reactivi-
ties, the potassium salts of the
imidazoles (2a–d)-K were not
isolated in substance, but were
generated by deprotonation of
the corresponding imidazoles
2ACTHNUTRGNE(UNG a–d)-H with KOtBu (typical-
ly 1.05 equivalents) in DMSO
in the flasks used for the kinet-
ic investigations.
Several combinations of the
heterocyclic anions 2–4 with
the benzhydrylium ions 1b–g
were also studied in water
(Table 3). Due to the low acidi-
ties of the heterocycles, aque-
ous solutions of their potassi-
Scheme 8. Products of the reactions of the heterocyclic anions 3 with the benzhydrylium ion 1b in DMSO at
RT.
130
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 127 – 137

Nucleophilic Reactivities
FULL PAPER
were able to realize conditions,
where the correction term did
not exceed 10% of kobs by
combining a large excess of the
heterocycles (2–4)-H with only
0.02 to 0.2 equivalents of
KOH. In few cases, where the
neutral species were almost in-
soluble in water, larger correc-
tions had to be made, as one
equivalent of KOH was
needed to give homogeneous
solutions. The intercepts of
these plots correspond to the
reactions of the electrophiles
with water and are generally
negligible in agreement with
previous work, where water
(N=5.20)[25] was demonstrated
Scheme 9. Products of the reactions of the pyrimidine anions 4 with the benzhydrylium ion 1b in DMSO.
to react much more slowly with benzhydrylium ions than
the nucleophiles investigated in this work.
Solvent effects: Figure 2 shows that the second-order rate
constants for the reactions of the theophylline anion 3d with
1e increases by a factor of 7 from almost pure water (con-
taining 3 vol% of DMSO) to 50 vol% aq. DMSO and by a
further factor of 300 in pure DMSO. As the corrections due
to the contribution of hydroxide (see Supporting Informa-
tion) are typically very small for 3d, the values in Figure 2
do not include a correction for hydroxide.
Figure 1. Plot of the absorbance A (635 nm) vs. time for the reaction of
1 f with the anion of 4-nitroimidazole (2e) in DMSO at 208C and corre-
lation of the first-order rate constants kobs values with the concentration
of 2e (insert).
um salts are partially hydrolyzed and contain hydroxide
anions. Therefore, three competing reactions may account
for the decay of the benzhydrylium ions in water. The ob-
served rate constants kobs for the consumption of the electro-
philes in water reflect the sum of the reactions with the
anionic nucleophiles 2–4 (k2), with hydroxide (k2,OH),[24] and
with water (kw) [Eq. (2)].
kobs ¼ k2½2 ꢁ 4ꢂ þ k2,OH½HOꢁꢂ þ kw
ð2Þ
ð3Þ
Figure 2. Second-order rate constant k2 of the reaction of the anion of
theophylline (3d) with 1e in mixtures of DMSO/water at 208C.
keff ¼ kobs ꢁ k2,OH½HOꢁꢂ ¼ k2½2 ꢁ 4ꢂ þ kw
The equilibrium concentrations of HOꢁ and 2–4 were cal-
culated from the initial concentrations of the reagents and
the pKaH values, as described in the Supporting Information.
Rearrangement of Equation (2), that is, subtraction of the
contribution of hydroxide (k2,OH from ref. [24]) from the ob-
served rate constant kobs, yields Equation (3). The second-
order rate constants k2 for the reactions of the benzhydryli-
um ions with 2–4 can then be obtained from plots of keff
versus the concentration of the nucleophiles. Usually, we
In line with this result, a large decrease of reactivity to-
wards the benzhydrylium ions 1a–g is also found for most
anions when DMSO is replaced by water as the solvent
(Tables 2 and 3, Figure 3). While the pyrimidine anions 4a
and 4b react approximately 10000 times more slowly in
water than in DMSO, a factor of only 500–5000 is found for
the azole anions 2 and 3. These effects can be rationalized
by the formation of hydrogen bonds of the anions 2–4 to-
wards water which reduces the negative charge and thereby
Chem. Eur. J. 2012, 18, 127 – 137
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
131

H. Mayr et al.
Table 2. Second-order rate constants k2 for the reactions of the nucleophiles 2–4 with the reference electro-
philes 1 in DMSO at 208C.
A remarkable difference can
be found in the behavior of the
anion of 1-methyluracil (4c) in
comparison to the other pyri-
midine anions, that is, 4a and
4b. While 4a and 4b react
only 13000 times slower in
water than in DMSO, a de-
crease of reactivity by a factor
of 250000 is found for the
anion of 1-methyluracil (4c).
Obviously, the imide anion 4c
is much better solvated in
water and is, therefore, signifi-
cantly less reactive in water
than the amide anions 4a and
4b, while the reactivity differ-
ences in DMSO are marginal.
Nu
N;
sN
Electro-
phile
k2
Nu
N;
sN
Electro-
phile
k2
[Lmolꢁ1 sꢁ1
]
[Lmolꢁ1 sꢁ1
]
1m
1l
1k
1j
1i
1h
1g
7.69ꢃ101
3.82ꢃ102
4.49ꢃ102
3.97ꢃ103
5.88ꢃ103
3.28ꢃ104
4.38ꢃ105
1k
1j
1i
1h
1g
1 f
6.88ꢃ101
4.91ꢃ102
9.32ꢃ102
6.05ꢃ103
9.17ꢃ104
2.11ꢃ105
19.13;
0.55
21.09;
0.51
1i
7.11ꢃ101
4.86ꢃ102
1.18ꢃ104
2.91ꢃ104
6.09ꢃ104
1.81ꢃ105
1m
1l
1k
7.83ꢃ101
4.54ꢃ102
5.28ꢃ102
5.25ꢃ102[a]
6.72ꢃ103
7.20ꢃ103
7.67ꢃ103[a]
4.03ꢃ104
4.21ꢃ105
1h
1g
1 f
1e
1d
16.29;
0.65
21.03;
0.50
1j
1i
1i
1.92ꢃ101
1.43ꢃ102
9.23ꢃ103
2.16ꢃ104
5.26ꢃ104
1.51ꢃ105
6.07ꢃ105
1h
1g
1h
1g
1 f
1e
1d
1c
15.03;
0.77
1m
1l
1k
1j
9.47ꢃ101
4.47ꢃ102
6.74ꢃ102
6.88ꢃ103
7.40ꢃ103
7.33ꢃ103[a]
4.04ꢃ104
3.98ꢃ104[a]
6.38ꢃ105
Correlation analysis: Linear
correlations were obtained,
when logk2 for the reactions of
the anionic nucleophiles 2–4
with the reference electro-
philes 1 were plotted against
the electrophilicity parameters
E, as shown for some represen-
tative examples in Figure 3. As
depicted in the Supporting In-
formation, all other reactions
investigated in this work fol-
lowed analogous linear correla-
tions indicating that Equa-
tion (1) is applicable to reac-
tions with these classes of nu-
cleophiles. The slopes of these
correlations equal the nucleo-
phile-specific parameters sN,
and the negative intercepts on
the abscissa (logk2 =0) corre-
spond to the nucleophilicity
parameters N.
21.29;
0.51
1i
1i
1.05ꢃ101
6.70ꢃ101
2.32ꢃ103
5.81ꢃ103
1.34ꢃ104
3.63ꢃ104
1.57ꢃ105
1h
1g
1 f
1e
1d
1c
1h
1g
14.78;
0.71
1m
1l
9.19ꢃ101
5.80ꢃ102
9.67ꢃ102
1.03ꢃ104
1.50ꢃ104
1.44ꢃ105
1k
1j
1i
1h
1g
1 f
1e
1.51ꢃ103
2.40ꢃ104
5.02ꢃ104
1.09ꢃ105
20.69;
0.60
18.00;
0.55
1h
1g
1 f
1e
1d
1c
1b
2.50ꢃ103
6.16ꢃ103
1.41ꢃ104
4.60ꢃ104
1.52ꢃ105
2.66ꢃ105
1i
2.15ꢃ102
9.73ꢃ102
3.66ꢃ104
7.30ꢃ104
1.56ꢃ105
3.43ꢃ105
1h
1g
1 f
1e
1d
14.81;
0.71
17.04;
0.63
1i
6.54ꢃ101
4.04ꢃ102
1.29ꢃ104
3.02ꢃ104
6.03ꢃ104
1.65ꢃ105
6.20ꢃ105
1i
4.12ꢃ102
2.20ꢃ103
4.94ꢃ104
1.05ꢃ105
2.44ꢃ105
6.93ꢃ105
1h
1g
1 f
1e
1d
1c
1h
1g
1 f
1e
1d
17.63;
0.62
16.06;
0.68
Structure–reactivity relation-
ships: The narrow range of sN
for all nucleophiles listed in
1i
9.71ꢃ101
7.88ꢃ102
3.18ꢃ104
5.32ꢃ104
1.83ꢃ105
3.25ꢃ105
1i
1.22ꢃ102
7.01ꢃ102
1.82ꢃ104
4.16ꢃ104
1.13ꢃ105
3.13ꢃ105
1.10ꢃ106
1h
1g
1 f
1e
1d
1h
1g
1 f
1e
1d
1c
16.36;
0.69
Tables 2 and 3 (0.50 < sN
<
16.40;
0.67
0.77), which is illustrated by
the almost parallel correlation
lines in Figure 3 implies that
the relative reactivities of
these anions depend only
slightly on the electrophilicity
of the reaction partners. The
[a]KOtBu was used as minor component (0.7 equiv).
the reactivity of the anionic nucleophile. Obviously, a rela-
tively small portion of water is sufficient for the formation
of hydrogen bonds and leads to a diminished nucleophilicity.
reactivities towards the benzhydrylium ion 1g, for which
most rate constants have directly been measured, therefore,
reflect general structure reactivity trends (Scheme 10).
132
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 127 – 137

Nucleophilic Reactivities
FULL PAPER
Table 3. Second-order rate constants for the reactions of 2–4 with the reference electrophiles 1 in water at 208C.
Nu
N;
sN
pKaH
Electro-
phile
k2/
Nu
N;
sN
pKaH
Electro-
phile
k2/
Lmolꢁ1 sꢁ1
Lmolꢁ1 sꢁ1
1g
1 f
1e
1d
4.53ꢃ100
1.27ꢃ101
1.98ꢃ101
4.75ꢃ101
1g
1 f
1e
1b
3.26ꢃ100
6.89ꢃ100
1.76ꢃ101
2.84ꢃ102
11.37;
0.53
10.77;
0.65
9.10[26]
9.80[33]
1g
1 f
1e
1d
3.27ꢃ100
6.21ꢃ100
1.49ꢃ101
2.55ꢃ101
1g
1 f
1e
1d
1c
1.41ꢃ101
1.73ꢃ101
5.74ꢃ101
9.48ꢃ101
2.07ꢃ102
11.07;
0.50
10.5[27]
8.57[28]
8.93[29]
12.09;
0.52
9.31[34]
9.45[35]
9.94[36]
1g
1e
1b
1a
1.42ꢃ101
5.32ꢃ101
7.16ꢃ102
1.56ꢃ104
1g
1 f
1e
1d
1c
2.71ꢃ100
4.27ꢃ100
1.08ꢃ101
2.11ꢃ101
4.54ꢃ101
11.52;
0.67
10.75;
0.53
1g
1 f
1e
1d
3.49ꢃ100
6.75ꢃ100
1.45ꢃ101
3.50ꢃ101
11.00;
0.54
1g
1 f
1e
1d
1c
3.86ꢃ100
6.92ꢃ100
1.80ꢃ101
2.89ꢃ101
6.03ꢃ101
11.17;
0.51
1g
1 f
1e
1d
1c
1.07ꢃ100[a]
2.69ꢃ100[a]
8.01ꢃ100[a]
2.09ꢃ101[a]
4.99ꢃ101[a]
10.06;
0.71
8.52[30]
1e
1d
1c
1b
7.26ꢃ10ꢁ1
1.41ꢃ100
5.56ꢃ100
1.39ꢃ101
8.54;
0.77
1g
1e
1d
1c
1b
3.43ꢃ100
1.93ꢃ101
4.13ꢃ101
1.06ꢃ102
2.40ꢃ102
9.99[37]
10.93;
0.62
9.80[31]
1e
1d
1c
1b
5.09ꢃ101
1.09ꢃ102
1.60ꢃ102
5.87ꢃ102
11.63;
0.59
9.21[32]
[a]In H2O/DMSO 95:5 (v/v).
The decreasing nucleophilicities of the imidazole anions 2
in the left column of Scheme 10 can be explained by a
better stabilization of the negative charge by electron-with-
drawing substituents, which is also reflected by the correla-
tion with the Hammett parameters sp,[36] depicted in
ring in 3a is replaced by the more electron-withdrawing pyr-
imidine ring in 3c. This effect is partially compensated by
the additional amino group in the adenine anion 3e which is
2.6 times more nucleophilic than the purine anion 3c. Anne-
lation of an uracil ring causes an even stronger reduction in
reactivity, and the anion of theophylline (3d) is even less re-
active than the 2- or 4-formyl substituted imidazole anions
2 f and 2g.
Methylation of the uracil anion (4a) at either N1 or C5
does not strongly affect its nucleophilicity (right column in
Scheme 10) and the anions of uracil (4a), thymine (4b), and
1-methyluracil (4c) are positioned between the anions of
benzimidazole (3a) and adenine (3e) in Scheme 10 (DMSO
part). While uracil and thymine are also equally reactive in
water, the 1-methyluracil anion (4c) is approximately 10
times less nucleophilic, although the additional methyl
group in 4c enhances its basicity in water by half an order
of magnitude. This can be rationalized by the different
mode of attack: While 4a and 4b react via the amide nitro-
ꢁ
Figure 4. Analogous correlations with sp and sm are shown
on page S5 of the Supporting Information.
A comparison of the reactivities of 2b and 2c as well as
of 2 f and 2g (Scheme 10) reveals that the position of the
substituent, that is, whether the substituent is in 2- or in 4-
position, is of minor relevance for the nucleophilicity of imi-
dazole anions.
From the reactivities of the anions of imidazole (2a), ben-
zimidazole (3a), and benzotriazole (3b) in DMSO one can
conclude that annelation of a benzene ring (2a ! 3a) re-
duces the nucleophilicity by a factor of 4.8 and furthermore
by a factor of 7.8 when an additional nitrogen is incorporat-
ed in the five-membered ring (3a ! 3b). A ten-fold de-
crease of reactivity is found when the annelated benzene
Chem. Eur. J. 2012, 18, 127 – 137
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
133

H. Mayr et al.
gen N1, this center is blocked
by the methyl group in 4c and
the nucleophilic attack occurs
at the imide nitrogen N3.
Due to their low solubility in
DMSO, even in the presence
of [18]crown-6, the guanine
salts 3ACHTNUTRGNEUNG(f–h)-K were only stud-
ied in water. In this solvent, all
compounds were found within
a narrow reactivity range of
one order of magnitude. The
guanine anion 3 f, where the
urea structure of theophylline
is replaced by
a guanidine
structure, is eight times more
reactive than 3d. A change
from the imidazole anion in
guanidine to the amide anion
in 9-methyl-guanine (3 f ! 3g)
goes along with a decrease of
reactivity by a factor of 2.6
showing that the nucleophilici-
ties of both ring systems are
comparable. Remarkably, an
Figure 3. Plots of the rate constants logk2 for the reactions of the heterocyclic anions 2–4 with reference elec-
trophiles 1 in DMSO and water versus their electrophilicity parameters E.
increase of reactivity by
a
factor of 4.3 is found, however,
Figure 4. Correlation of the gross reactivities (logk2 for the reactions of
different 4-substituted imidazole anions 2 with the benzhydrylium ion 1g
in DMSO at 208C) with Hammettꢄs substitution constants sp.
when the methyl group is exchanged by ribose in the anion
of guanosine (3h).
Comparison with neutral nucleophiles: For practical reasons,
which have previously been specified, the kinetic investiga-
tions of the neutral imidazole and benzimidazole had to be
performed in either DMSO or acetonitrile.[21] It has been
shown, however, that the rate constants for the reactions of
the parent imidazole with several benzhydrylium ions dif-
fered by less than a factor of 1.4 in these two solvents.
Therefore, rate constants for neutral azoles in acetonitrile
have been employed for the comparison in Scheme 11 when
data in DMSO were not available. Scheme 11 shows that in
all systems compared, the anions are 104–105 times more nu-
cleophilic than their neutral counterparts.
Scheme 10. Comparison of the gross reactivities (at 208C) of azole anions
with the benzhydrylium ion 1g in DMSO and water; entries for 3 f and
4c in water were calculated using Eq. (1).
134
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 127 – 137

Nucleophilic Reactivities
FULL PAPER
activity towards the benzhydrylium ion 1g and Brønsted ba-
sicity in DMSO is remarkably good whereas no relation be-
tween basicity and nucleophilicity is found in water. Though
only few pKaH values for the anions 2–4 are available in
DMSO, one can derive that the slope obtained for the hetero-
cyclic anions 2–4 (0.23) is slightly smaller than that obtained
for other amide and imide anions,[18a] indicating that also in
DMSO variation of the basicity of these anions has little in-
fluence on their nucleophilicity.
Conclusion
The rate constants for the reactions of imidazole, purine,
and pyrimidine anions with quinone methides and benzhy-
drylium ions follow the linear free-energy relationship [Eq.
(1)] allowing us to include these compounds into our com-
prehensive nucleophilicity scales and compare their nucleo-
philicities with those of other nucleophiles (Figure 6). Be-
cause of the different sN values of these anions, which imply
that the relative reactivities of the nucleophiles depend
slightly on the reactivity of the electrophilic reaction part-
ner, the ordering of the nucleophiles in Figure 6 deviates
somewhat from that in Scheme 10.
In DMSO, these heterocyclic anions span over more than
six orders of reactivity and are comparable to carbanions,
amide and imide anions, or amines; in water, a smaller
range of reactivity is observed. The poor correlation be-
tween Brønsted basicity and nucleophilicity in water shows
that pKaH values cannot be used for the prediction of rela-
tive reactivities. This deviation may be due to the fact that
pKaH values refer to reactions with the proton, while the nu-
cleophilicity parameters N refer to reactions with carbon
electrophiles.
For unsymmetrically substituted azole anions, it obviously
depends on the structure of the anion whether more than
one regioisomer can be isolated. On the other hand, the
uracil anions 4 are selectively attacked at nitrogen by the
studied electrophiles, and O-alkylation has not been ob-
served.
Scheme 11. Comparison of the second-order rate constants (krel) of the
reaction of the benzhydrylium ion 1g with the anions 2 or 3 and with
their neutral analogues 2-H or 3-H (in DMSO, from ref. [21]). [a] k2
value was calculated using previously published N and sN parameters
from ref. [21] and EACHTUNGTRENNUNG(1g) from Table 1. [b] Data for the neutral analogue
refers to CH3CN.
Correlation with Brønsted basicities: Brønsted basicity is
often used as a guide to estimate the nucleophilic reactivity
despite the poor quality that was often observed in these
correlations. Figure 5 shows that the correlation between re-
The results presented in this paper can be used together
with the results of ongoing research on the leaving group
abilities of these heterocyclic anions as well as of their neu-
tral counterparts to get a better understanding of DNA al-
kylation and the related repair processes.
Experimental Section
Synthesis of the potassium salts (2–4)-K: The NH-acid was added to a so-
lution of KOtBu in dry ethanol or to a solution of KOH in water and the
mixture was stirred for 10 min. Then, the solvent was evaporated under
reduced pressure and the solid residue was washed several times with dry
diethyl ether and filtrated under N2. The benzimidazole-potassium salt
3b-K was prepared as already described for the sodium salt.[39] 1-Methyl-
uracil was synthesized according to ref. [40].
Figure 5. Relationship between Brønsted basicity and logk2 for the reac-
tion of 1g with several heterocyclic nucleophiles in DMSO (*) and
water (*). [pKaH in DMSO: 2a: 18.6 (ref. [37]), 2b: 19.9 (ref. [38]), 3a:
16.4 (ref. [37]), 3b: 12.6 (ref. [38]), 3e: 14.2 (ref. [37]), 4a: 14.1 (ref. [37]);
pKaH in water: see Table 3].
Reactions of the potassium salts (2–4)-K with quinone methides 1k and
1l: The potassium salts (2–4)-K (or the corresponding azole and
Chem. Eur. J. 2012, 18, 127 – 137
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
135

H. Mayr et al.
Photophysics
ters
(Applied
SX.18MV-R or Hi-Tech SF-61DX2)
were used for the investigation of
rapid reactions (t1/2 <10 s). The tem-
perature of all solutions was kept
constant at 20.0ꢃ0.18C during all ki-
netic studies by using a circulating
bath thermostat. In all runs the nucle-
ophile concentration was at least 10
times higher than the concentration
of the electrophile, resulting in
pseudo-first-order kinetics with an ex-
ponential decay of the electrophileꢄs
concentration. First-order rate con-
stants kobs (sꢁ1
) were obtained by
least-squares fitting of the absorbance
data to a single-exponential At =A0
exp
(ꢁkobst)
+
C. The second-order
were
rate constants k2 (Lmolꢁ1 sꢁ1
)
obtained from the slopes of the linear
plots of kobs against the nucleophileꢄs
concentration.
Acknowledgements
We thank the Fonds der Chemischen
Industrie (scholarship to M.B.) and
the Deutsche Forschungsgemeinschaft
(SFB 749) for financial support. We
are grateful to Dr. Frank Brotzel for
experimental support and Dr. Armin
Ofial and Dr. Roland Appel for help-
ful discussions.
[1] a) A. F. Pozharskii, A. D. Gar-
novskii, A. M. Simonov, Russ.
Chem. Rev. 1966, 35, 122–144;
b) M. R. Grimmett in Compre-
Figure 6. Comparison of the nucleophilicities N of heterocyclic anions with other C- and N nucleophiles in
DMSO (nucleophiles determined in other solvents are marked).[19f]
hensive
Organic
Chemistry,
1.05 equiv KOtBu) were dissolved in dry DMSO (ca. 5 mL) and a solu-
tion of the electrophile in DMSO (ca. 5 mL, each with ca. 5–10% CH2Cl2
Vol. 4 (Eds.: D. Barton, W. D.
Ollis), Pergamon, Oxford, 1979,
as co-solvent) was added. The mixture was stirred for 15 min before
0.5% acetic acid (ca 50 mL) was added. The mixture was extracted with
dichloromethane or ethyl acetate (3ꢃ40 mL) and the combined organic
phases were washed with saturated NaCl solution (3ꢃ40 mL), dried over
Na2SO4, and evaporated under reduced pressure. The crude reaction
products were purified by column chromatography on Al2O3 (activity
grade iv) and subsequently characterized by NMR spectroscopy and MS.
pp. 357–410; c) M. R. Grimmett, Adv. Heterocycl. Chem. 1980, 27,
241–326; d) A. F. Pozharskii, A. T. Soldatenkov, A. R. Katritzky,
Heterocycles in Life and Society, 2nd ed., Wiley, Chichester, 2011;
e) A. R. Katritzky, A. F. Pozharskii, Handbook of Heterocyclic
Chemistry, Pergamon, Oxford, 2000; f) A. R. Katritzky, X. Lan, J. Z.
Yang, O. V. Denisko, Chem. Rev. 1998, 98, 409–548.
[2] F. Schneider, Angew. Chem. 1978, 90, 616–625; Angew. Chem. Int.
Ed. Engl. 1978, 17, 583–592.
[3] H. A. Staab, Angew. Chem. 1962, 74, 407–423; Angew. Chem. Int.
Ed. Engl. 1962, 1, 351–367.
[4] a) A. Grimison, J. H. Ridd, B. V. Smith, J. Chem. Soc. 1960, 1352–
1356; b) J. H. Ridd, B. V. Smith, J. Chem. Soc. 1960, 1363–1369.
[5] M. W. Miller, H. L. Howes, R. V. Kasubick, A. R. English, J. Med.
Chem. 1970, 13, 849–852.
[6] a) F. Terrier, F. Debleds, J. C. Halle, M. P. Simonnin, Tetrahedron
Lett. 1982, 23, 4079–4082; b) M. P. Simonnin, J. C. Halle, F. Terrier,
M. J. Pouet, Can. J. Chem. 1985, 63, 866–870.
[7] a) W. F. Veldhuyzen, Y.-F. Lam, S. E. Rokita, Chem. Res. Toxicol.
2001, 14, 1345–1351; b) W. F. Veldhuyzen, A. J. Shallop, R. A. Jones,
S. E. Rokita, J. Am. Chem. Soc. 2001, 123, 11126–11132; c) M. Frec-
cero, R. Gandolfi, M. Sarzi-Amadꢅ, J. Org. Chem. 2003, 68, 6411–
6423; d) E. E. Weinert, K. N. Frankenfield, S. E. Rokita, Chem. Res.
Toxicol. 2005, 18, 1364–1370; e) E. E. Weinert, R. Dondi, S. Collor-
Reactions of the potassium salts (2–4)-K with the benzhydrylium tetra-
fluoroborate 1b-BF4: The potassium salts (2–4)-K (or the corresponding
azole and 1.05 equiv KOtBu) were dissolved in dry DMSO (ca 5 mL) and
a solution of 1b-BF4 in DMSO (ca 5 mL) was added. After stirring for
15 min, the mixture was diluted with ethyl acetate (ca 40 mL). DMSO
was removed by washing several times with saturated NaCl solution (ca
30 mL each) and ethyl acetate was evaporated under reduced pressure.
The reaction products were subsequently characterized by NMR.
Kinetics: As the reactions of colored benzhydrylium ions or quinone me-
thides with colorless nucleophiles result in colorless products, the reac-
tions could be followed by UV/Vis spectroscopy. Slow reactions (t1/2
>
10 s) were monitored by using a conventional diode array UV/Vis spec-
trophotometer (J&M TIDAS) that was connected to a Hellma 661.502-
QX quartz Suprasil immersion probe (5 mm light path) by fiber optic
cables with standard SMA connectors. Stopped-flow UV/Vis spectrome-
136
www.chemeurj.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2012, 18, 127 – 137

Nucleophilic Reactivities
FULL PAPER
edo-Melz, K. N. Frankenfield, C. H. Mitchell, M. Freccero, S. E.
Rokita, J. Am. Chem. Soc. 2006, 128, 11940–11947.
[22] S. Hoz, H. Basch, J. L. Wolk, T. Hoz, E. Rozental, J. Am. Chem.
Soc. 1999, 121, 7724–7725.
[8] J. C. Martin, D. F. Smee, J. P. H. Verheyden, J. Org. Chem. 1985, 50,
755–759.
[23] a) D. Shugar, J. J. Fox, Biochim. Biophys. Acta 1952, 9, 199–218;
b) K. Nakanishi, N. Suzuki, F. Yamazaki, Bull. Chem. Soc. Jpn. 1961,
34, 53–57; c) E. Wittenburg, Chem. Ber. 1966, 99, 2391–2398; d) S.
Ganguly, K. K. Kundu, Can. J. Chem. 1994, 72, 1120–1126; e) M. A.
Kurinovich, J. K. Lee, J. Am. Soc. Mass Spectrom. 2002, 13, 985–
995; f) A. K. Chandra, T. Uchimaru, T. Zeegers-Huyskens, J. Mol.
Struct. 2002, 605, 213–220.
[24] S. Minegishi, H. Mayr, J. Am. Chem. Soc. 2003, 125, 286–295.
[25] S. Minegishi, S. Kobayashi, H. Mayr, J. Am. Chem. Soc. 2004, 126,
5174–5181.
[9] R. V. Joshi, J. Zemlicka, Tetrahedron 1993, 49, 2353–2360.
[10] S. Boncel, A. Gondela, K. Walczak, Synthesis 2010, 1573–1589.
[11] A. Copik, J. Suwinski, K. Walczak, J. Bronikowska, Z. Czuba, W.
Krꢆl, Nucleosides Nucleotides Nucleic Acids 2002, 21, 377–383.
[12] O. D. Gupta, B. Twamley, R. L. Kirchmeier, J. M. Shreeve, J. Fluo-
rine Chem. 2000, 106, 199–204.
[13] D. E. Jane, K. Hoo, R. Kamboj, M. Deverill, D. Bleakman, A. Man-
delzys, J. Med. Chem. 1997, 40, 3645–3650.
[14] H. J. Cleaves II, Astrobiology 2002, 2, 403–415.
[26] T. C. Bruice, G. L. Schmir, J. Am. Chem. Soc. 1958, 80, 148–156.
[27] M. G. Elliott, R. E. Shepherd, Inorg. Chem. 1987, 26, 2067–2073.
[28] A. Albert, R. J. Goldacre, J. Phillips, J. Chem. Soc. 1948, 2240–2249.
[29] A. Albert, D. J. Brown, J. Chem. Soc. 1954, 2060–2071.
[30] G. Vçlgyi, R. Ruiz, K. Box, J. Comer, E. Bosch, K. Takꢇcs-Novꢇk,
Anal. Chim. Acta 2007, 583, 418–428.
[15] a) A. Gambacorta, M. E. Farah, D. Tofani, Tetrahedron 1999, 55,
12615–12628; b) A. Gambacorta, D. Tofani, M. A. Loreto, T. Gas-
peri, R. Bernini, Tetrahedron 2006, 62, 6848–6854.
[16] a) R. G. Pearson, J. Am. Chem. Soc. 1963, 85, 3533–3539; b) R. G.
Pearson, Science 1966, 151, 172–177; c) R. G. Pearson, J. Songstad,
J. Am. Chem. Soc. 1967, 89, 1827–1836.
[31] H. F. W. Taylor, J. Chem. Soc. 1948, 765–766.
[17] a) G. Klopman, J. Am. Chem. Soc. 1968, 90, 223–234; b) L. Salem,
J. Am. Chem. Soc. 1968, 90, 543–552.
[32] H. A. Azab, Z. M. Anwar, M. Sokar, J. Chem. Eng. Data 2004, 49,
256–261.
[18] a) M. Breugst, T. Tokuyasu, H. Mayr, J. Org. Chem. 2010, 75, 5250–
5258; b) M. Breugst, H. Mayr, J. Am. Chem. Soc. 2010, 132, 15380–
15389; c) H. Mayr, M. Breugst, A. R. Ofial, Angew. Chem. 2011,
123, 6598–6634; Angew. Chem. Int. Ed. 2011, 50, 6470–6505.
[19] a) H. Mayr, T. Bug, M. F. Gotta, N. Hering, B. Irrgang, B. Janker, B.
Kempf, R. Loos, A. R. Ofial, G. Remennikov, H. Schimmel, J. Am.
Chem. Soc. 2001, 123, 9500–9512; b) R. Lucius, R. Loos, H. Mayr,
Angew. Chem. 2002, 114, 97–102; Angew. Chem. Int. Ed. 2002, 41,
91–95; c) H. Mayr, B. Kempf, A. R. Ofial, Acc. Chem. Res. 2003, 36,
66–77; d) H. Mayr, A. R. Ofial, Pure Appl. Chem. 2005, 77, 1807–
1821; e) D. Richter, N. Hampel, T. Singer, A. R. Ofial, H. Mayr,
Eur. J. Org. Chem. 2009, 3203–3211; f) For a comprehensive listing
of nucleophilicity parameters N, sN and electrophilicity parameters
E, see http://www.cup.uni-muenchen.de/oc/mayr/DBintro.html.
[20] H. Mayr, M. Patz, Angew. Chem. 1994, 106, 990–1010; Angew.
Chem. Int. Ed. Engl. 1994, 33, 938–957.
[33] W. Pfleiderer, Liebigs Ann. Chem. 1961, 647, 167–173.
[34] A. Albert, Biochem. J. 1953, 54, 646–654.
[35] P. A. Levene, L. W. Bass, H. S. Simms, J. Biol. Chem. 1926, 70, 229–
241.
[36] C. Hansch, A. Leo, D. Hoekman, Exploring QSAR - Hydrophobic,
Electronic, and Steric Constants, American Chemical Society, Wash-
ington, DC, 1995; and references therein.
[37] F. G. Bordwell, Acc. Chem. Res. 1988, 21, 456–463.
[38] M. I. Terekhova, E. S. Petrov, E. M. Rokhlina, D. N. Kravtsov, A. I.
Shatenshtein, Khim. Geterotsikl. Soedin. 1979, 1104–1108; Chem.
Heterocycl. Compd. 1979, 904–908.
[39] E. Anders, J. G. Tropsch, A. R. Katritzky, D. Rasala, J. J. Vanden
Eynde, J. Org. Chem. 1989, 54, 4808–4812.
[40] W. Micklitz, B. Lippert, H. Schoellhorn, U. Thewalt, J. Heterocycl.
Chem. 1989, 26, 1499–1500.
[21] M. Baidya, F. Brotzel, H. Mayr, Org. Biomol. Chem. 2010, 8, 1929–
1935.
Received: August 3, 2011
Published online: December 8, 2011
Chem. Eur. J. 2012, 18, 127 – 137
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
137