Controlling the outcome of an N-alkylation reaction by using N-oxide functional groups

Article abstract of DOI:10.1021/jo0503106

Covalent modifiers of proteins are of importance in chemical proteomics, an emerging chemical technology used to assign protein function. In this study, high-field ¹H NMR techniques were used to analyze the reaction of the bioactive compound, 2,3-bis(bromomethyl)quinoxaline 1,4-dioxide, with amines (a model system for proteins containing nitrogen-based nucleophiles). Unexpectedly, the results show that a double nucleophilic substitution reaction involving 2 equiv of the amine is preferred to an intramolecular cyclization pathway. A direct comparison with the reaction carried out on a substrate lacking the N-oxide functional groups is also provided. X-ray crystal structures and computational studies are used to rationalize the observed differences in reactivity between the two systems.

Full text of DOI:10.1021/jo0503106

Controlling the Outcome of an N-Alkylation Reaction by Using
N-Oxide Functional Groups
Russell J. Pearson, Kathryn M. Evans, Alexandra M. Z. Slawin, Douglas Philp, and
Nicholas J. Westwood*
Centre for Biomolecular Sciences, School of Chemistry, University of St. Andrews, North Haugh, St.
Andrews KY16 9ST, United Kingdom
njw3@st-andrews.ac.uk.
Received February 18, 2005
Covalent modifiers of proteins are of importance in chemical proteomics, an emerging chemical
technology used to assign protein function. In this study, high-field 1H NMR techniques were used
to analyze the reaction of the bioactive compound, 2,3-bis(bromomethyl)quinoxaline 1,4-dioxide,
with amines (a model system for proteins containing nitrogen-based nucleophiles). Unexpectedly,
the results show that a double nucleophilic substitution reaction involving 2 equiv of the amine is
preferred to an intramolecular cyclization pathway. A direct comparison with the reaction carried
out on a substrate lacking the N-oxide functional groups is also provided. X-ray crystal structures
and computational studies are used to rationalize the observed differences in reactivity between
the two systems.
Introduction
reaction of 1 with primary amines was more complex
than first envisaged. This reaction has been the subject
of previous reports,5-7 culminating in the identification
of 5 (R ) CH3) as the major compound produced on
treatment of 1 with methylamine at low temperature.6,8
This transformation was proposed to occur by the mech-
anism shown in path A of Scheme 1, although no evidence
in support of this sequence was provided. In brief, it was
proposed that two molecules of methylamine react with
1 to give 3 (R ) CH3). An intramolecular reaction
The development of novel chemical probes that co-
valently modify their protein target(s) is of considerable
importance in chemical proteomics.1-4 In the majority of
cases, the mechanism of covalent protein modification is
obvious. However, during recent attempts to develop a
novel chemical probe based on 2,3-bis(bromomethyl)-
quinoxaline 1,4-dioxide (1), it became clear that the
* Address correspondence to this author. Phone: +44-1334-463816.
Fax: +44-1334-463808.
(1) Vocadlo, D. J.; Bertozzi, C. R. Angew. Chem., Int. Ed. 2004, 43,
(5) Landquist, J. K.; Silk, J. A. J. Chem. Soc. 1956, 2052-2063.
(6) Anderson, R. C.; Fleming, R. H. Tetrahedron Lett. 1969, 1581-
1584.
(7) Ried, W.; Grabosch J. Chem. Ber. 1958, 2485-2495.
(8) Katritzky, A. R.; Lagowski, J. M. In Chemistry of the heterocyclic
N-oxides, 1st ed.; Chapter 4, Reactions of substituents on heterocyclic
N-oxides; Blomquist, A. T., Ed.; Academic Press: New York, 1971; p
389.
5338-5342.
(2) Campbell, D. A.; Szadenings, A. K. Curr. Opin. Chem. Biol. 2003,
7, 296-303.
(3) Jeffery, D. A.; Bogyo, M. Curr. Opin. Chem. Biol. 2003, 14, 87-
95.
(4) Kidd, D.; Liu, Y.; Cravatt, B. F. Biochemistry 2001, 40, 4005-
4015.
10.1021/jo0503106 CCC: $30.25 © 2005 American Chemical Society
Published on Web 05/19/2005
J. Org. Chem. 2005, 70, 5055-5061
5055

Pearson et al.
SCHEME 1. Two Possible Reaction Pathways for
Conversion of Compound 1 to 5a
FIGURE 1. Relative abundance of compounds 7, 9, and 10.15a
SCHEME 2. Products Formed on Reaction of 7
with n-Butylaminea
a Reaction conditions: (i) methylamine, -80 °C; (ii) n-butyl-
amine, 25 °C.
involving the N-oxide functionality of the tautomer of 3
was then proposed to give the corresponding imine 4 (R
) CH3) en route to 5 (R ) CH3).
a Reaction conditions: (i) n-butylamine, 25 °C.15a
An analogue of 1 with substituents at the C(6) and C(7)
positions exhibits antibacterial activity,9,10 and 1 ir-
reversibly inhibits host cell invasion by the parasite
Toxoplasma gondii.11 The mode of action of these com-
pounds remains unassigned but almost certainly involves
alkylation of nucleophilic residues present in one or more
proteins. The reaction of 1 with amines is therefore of
importance in explaining the details of the chemical
interaction of 1 with potential protein targets. Here we
present detailed 1H NMR studies on the reaction of 1 with
primary (and secondary) amines. During the course of
these studies, several alternative mechanisms for the
conversion of 1 to 5a were considered. These included a
chemically intuitive route involving the formation of 6a
through a cyclization reaction (Scheme 1, path B).12
1
studied with use of 500 MHz H NMR spectroscopy.15a
Spectra were acquired at regular time points over a 17-h
period and the signals corresponding to the starting
material 7, major reaction intermediate 8, and products
9, 10, and 11 were tracked.
The 1H NMR studies showed that 7 reacts with
n-butylamine to give two major products, 9 and 10
(Scheme 2 and Figure 1).16 The ratio of 9:10 after 17 h
was 3:2. During the course of the reaction, minor signals
corresponding to the formation of 11 (Scheme 2) were also
observed due to decomposition of 9.17
Compound 9 is formed by cyclization of 8 (Scheme 3)
although this step must be relatively slow, resulting in
the formation of 10. H NMR studies suggest that 10 is
1
formed by an intermolecular reaction of 8 with 7 followed
by incorporation of a second molecule of n-butylamine
and macrocyclization.16 As expected, repeating the reac-
tion with a 4 mM initial concentration of 7 (a 5-fold
dilution)15a increased the ratio of 9:10 to 7:2 further in
Results and Discussion
Reaction of 2,3-Bis(bromomethyl)quinoxaline (7)
with Primary and Secondary Amines. 2,3-Bis(bro-
momethyl)quinoxaline (7) was prepared from commer-
cially available 1,2-diaminobenzene and 1,4-dibromobu-
tanedione.13,14 The reaction of 7 with n-butylamine was
(15) Standard reaction conditions consisted of (a) substrate (initial
concentration 20 or 4 mM) and n-butylamine (60 or 20 mM) in CDCl3
at 25 °C studied with 500 MHz 1H NMR spectroscopy, or (b) substrate
(initial concentration 20 mM) and diethylamine (60 mM) in CDCl3 at
25 °C, studied with 300 MHz 1H NMR spectroscopy.
(9) Glushkov, R. G.; Vozyakova, T. I.; Adamskaya, E. V.; Aleinikova,
S. A.; Radkevich, T. P.; Shepilova, L. D.; Padeiskaya, E. N.; Gus’kova,
T. A. Pharm. Chem. J. (Engl. Transl.) 1994, 28, 17-20.
(10) Similar compounds have also been shown to exhibit antibacte-
rial activity. McIlwain, H. J. Chem. Soc. 1943, 322-325. Kim, H. K.;
Miller, L. F.; Bambury, R. E.; Ritter, H. W. J. Med. Chem. 1977, 20,
557-560.
(11) Carey, K. L.; Westwood, N. J.; Mitchison, T. J.; Ward, G. E.
Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 7434-7438.
(12) Shields, J. E.; Bornstein, J. Chem. Ind. 1967, 1404-1405.
(13) Moriconi, E. J.; Fritsch, A. J. J. Org. Chem. 1965, 30, 1542-
1547.
(14) Deady, L. W.; Desneves, J.; Ross, A. C. Tetrahedron 1993,
9823-9828.
(16) A compound of this general structure has been previously
reported as a minor component of a reaction: Kreher, R.; Use, G.
Tetrahedron Lett. 1978, 4671-4674. The formation of 10 probably
occurs by reaction of 8 with 7. Reaction of 8 with itself is considered
unlikely, at least during the early stages of the reaction, due to the
relatively low concentration of 8 present compared with 7. A set of 1H
NMR signals (singlet at 4.86 ppm) corresponding to an additional
minor intermediate containing a CH2Br functional group was observed.
Attempts to assign the structure of this intermediate further proved
unsuccessful. An alternative route to 10 involving the reaction of 12
with 7 was ruled out due to the absence of signals corresponding to
12.
(17) Ried, W.; Grabosch J. Chem. Ber. 1958, 2485-2495.
5056 J. Org. Chem., Vol. 70, No. 13, 2005

Controlling the Outcome of an N-Alkylation Reaction
SCHEME 3. A Summary of Reaction Pathwaysa
Available to Intermediate 815a
a Evidence in support of the formation of 9 (major) and 10
(minor) was obtained. 1H NMR signals corresponding to the
formation of 12 were not observed.
SCHEME 4. Reaction of 7 with a Secondary
Aminea
FIGURE 2. Relative abundance of 1 and 5a following reaction
of 1 with n-butylamine.15a
a Reaction conditions: (i) diethylamine, 25 °C.15b Compound 14
is assigned as the salt based on symmetrical 1H NMR signals and
a shift in the benzylic methylene (singlet at 5.55 ppm, cf. singlet
at 4.35 ppm for the benzylic methylene in compound 17, Scheme
6).
favor of the intramolecular reaction (cf. 3:2 at 20 mM).
Interestingly, no signals corresponding to the formation
of the disubstituted compound, 12 (Scheme 3), were
observed. Signals corresponding to the formation of 8
were transiently observed,18 although isolation of 8 was
not possible, as predicted.
The situation is simplified further when 7 is reacted
with the secondary amine, diethylamine, under analo-
gous conditions.15b This results almost exclusively in the
formation of 14 (>90% as judged by 1H NMR analysis of
the reaction mixture, 74% isolated yield, Scheme 4).18 In
contrast to the reaction of 7 with n-butylamine, a 1H
NMR spectrum of this reaction taken after 1 h shows no
signals corresponding to the presence of 7 [cf. time
interval at 4346 s (1.2 h) in Figure 1]. The expected
increase in rate of the first nucleophilic displacement (for
a secondary versus primary amine) coupled with a
proposed reduction in the rate of the reaction of 13 with
any remaining 7 (cf. reaction of 8 with 7) on steric
grounds probably explains the high efficiency of the
conversion of 7 to 14.
The observation that intermediates 8 and 13 prefer to
undergo a cyclization reaction raises questions about the
proposed mechanism of formation of 5a from 1 (Scheme
1, path A).6 For example, if the proposed mechanism is
correct, what factors contribute to the differential reac-
tivity of intermediate 2a compared with intermediate 8?
It was therefore decided to study the reaction of 2,3-bis-
(bromomethyl)quinoxaline 1,4-dioxide (1) with n-butyl-
amine with the same approach.
FIGURE 3. Tracking key intermediates 2a and 3a together
with the major product 5a from reaction of 1 with n-butyl-
amine.15a
n-butylamine.15a The data show the disappearance of the
1H NMR signals corresponding to 1 with concomitant
formation of those corresponding to 5a. When carried out
on a preparative scale 5a was isolated following rapid
column chromatography in 90% yield.
Signals corresponding to several intermediates were
also observed, but a detailed analysis became possible
only when the experiment was repeated at a 5-fold
dilution (Figure 3).15a The observed formation of signals
corresponding to 2a followed by the delayed formation
of 3a is consistent with the majority of 5a being formed
by path A in Scheme 1. The fact that the 1H NMR signals
corresponding to formation of 3a increase while those
corresponding to the presence of 2a are decreasing is also
supportive of the proposed reaction mechanism. Integra-
1
tion and summation of key H NMR signals associated
with compounds 1, 2a, 3a, and 5a at intervals throughout
the time course indicate that these four compounds make
up >90% of the reaction mixture at all times.18
Reaction of 2,3-Bis(bromomethyl)quinoxaline 1,4-
Dioxide (1) with Primary and Secondary Amines.
Compound 1 was synthesized by two different literature
protocols, the more robust proving to be N-oxidation of
7, using purified mCPBA.5,19 Figure 2 shows a simplified
representation of the analysis of the reaction of 1 with
1
The H NMR spectra of 2a and 3a differ considerably
in the aromatic region, where 2a has four signals and
3a has two signals consistent with their respective
symmetry. In addition, 2a also has a signal at 4.99 ppm
for the CH2Br, which is absent in 3a (singlet 4.28 ppm,
benzylic methylenes).18
(18) For further details including detailed spectroscopic information
see the Supporting Information.
(19) Haddadin, M. J.; Samaha, M. S.; Hajj-Ubayd, A. B. Heterocycles
1992, 33, 541-544.
These studies therefore provide the first direct evidence
for the mechanism of this reaction (Scheme 1, path A).6
J. Org. Chem, Vol. 70, No. 13, 2005 5057

Pearson et al.
SCHEME 5. A Summary of Reaction Pathwaysa
Available to Intermediate 2a15a
a Evidence in support of the preferential formation of 3a was
obtained. No evidence in support of the formation of 6a was
obtained (Scheme 1, path B).18
SCHEME 6. Reaction of 1 with a Secondary
Aminea
FIGURE 4. X-ray structures of compounds 1 (top) and 7
(bottom).
initially favorable SN2 displacement would result in the
formation of 2a with both the secondary amine and the
remaining bromine functionalities on the same face. A
rotation about the C(2)-C(9) and/or C(3)-C(10) bond
must then occur as the transition state for cyclization is
approached (see Figure 5, part iii, for the atom number-
a Reaction conditions: (i) diethylamine, 25 °C.15b A singlet at
4.35 ppm was observed for the benzylic methylene in compound
17 (cf. the analogous CH2 in 14, 5.55 ppm, Scheme 4).
They also rule out several alternative mechanisms for
the transformation of 1 to 5a. In particular, no evidence
to support a reaction pathway that proceeds via inter-
mediate 6a was obtained (Scheme 1, path B).18 Minor
signals corresponding to the previously unreported 15
were observed (singlet, 4.71 ppm, benzylic methylenes).
Unexpectedly, these studies show that intermediate 2a
prefers to react with a second molecule of n-butylamine
to form 3a (Scheme 5), in contrast to the situation with
intermediate 8, which undergoes a cyclization reaction
(Scheme 3). This is despite the fact that the only
structural difference between these two intermediates is
the presence of the N-oxide functionalities.18 The reac-
tions of 1 and 7 with a secondary amine emphasize
further this differential reactivity. In contrast to 7, which
reacts to give the cyclized product 14 (Scheme 4), 1 reacts
with diethylamine under analogous conditions to give the
disubstituted product 17 (Scheme 6) as the only identifi-
able product (>90% yield as judged by 1H NMR analysis
of the reaction mixture).18
1
ing). Repetition of these H NMR studies in d6-DMSO
showed that the major products and intermediates were
as observed in CDCl3 and no additional signals corre-
sponding to the formation of 6a were present. This
indicates that hydrogen bonding interactions capable of
restricting the required bond rotations are not important
in determining the outcome of this reaction (data not
shown).20
Computational studies were carried out to help visual-
ize more effectively the relevant transition states (TS,
Figure 5(i)). In both cases, a TS model consistent with
an SN2-like cyclization reaction was obtained. However,
to achieve an attack angle for the approaching secondary
amine of greater than 160°, significant puckering of the
aromatic skeleton itself is required in 2a‡. The aromatic
backbone of 8‡ is essentially planar. The dihedral angle
Br(9)-C(9)-C(2)-N also provides a clear readout of the
differences between the two systems (Figure 5(iii)). In
2a‡, the bromine atom sits 57.8° out of the plane of the
aromatic backbone, compared to only 29.6° in the case
of 8‡. The electrostatic potential surface maps for 2a‡ and
8‡ (Figure 5(ii)) show that the difference in these dihedral
angles results from the buildup of negative charge on the
bromine atom in 2a (late transition state). As a result,
the transition state for cyclization of 2a is disfavored due
to electrostatic repulsion between the N-oxide oxygen
atom and the bromine.21 Further evidence in support of
the view that 2a‡ is considerably more distorted than 8‡
(and hence less energetically accessible) comes from the
difference in the calculated dihedral angle N(1)-C(2)-
C(3)-N(4) (-8.8 in 2a‡ and -2.6 in 8‡). The calculated
dihedral angles C(9)-C(2)-C(3)-C(10) and N-C(9)-
Several possible explanations were considered to ra-
tionalize the observed differences. Two of these centered
on the possibility that the secondary nitrogen atom
present in 2a cannot act as a nucleophile due to either
its protonation state or the fact that it is rapidly
converted to the corresponding imine. The first possibility
was ruled out based on the observed chemical shifts of
the benzylic methylenes in 2a (4.27 ppm for 2a; cf. 5.55
1
ppm for 14, Scheme 4), the second, by H NMR studies
that show that imine production from these types of
substrates is relatively slow (data not shown).18
X-ray crystallographic analysis of 1 and 7 (Figure 4)
showed that, in the solid state, the two bromine atoms
are located almost perpendicular to the plane of the
heterocyclic ring system and on opposite faces. Assuming
that this is the preferred conformation in solution, the
(20) Hunter, C. A. Angew. Chem., Int. Ed. 2004, 43, 5310-5324.
(21) Raposo, C.; Wilcox, C. S. Tetrahedron Lett. 1999, 40, 1285-
1288.
5058 J. Org. Chem., Vol. 70, No. 13, 2005

Controlling the Outcome of an N-Alkylation Reaction
SCHEME 7. Reaction of 18 with a Secondary
Aminea
Reaction conditions: (i) diethylamine, 25 °C.15b The ratio of 19:
20 was calculated as 63:37 in favor of the intramolecular process.
functional group leads to product 19 via cyclization and
(ii) initial displacement of the alternative bromine atom
leads to the disubstituted product 20. To test this
hypothesis, 18 was synthesized from 7 by mono-N-
oxidation, using 1 equiv of mCPBA. 1H NMR analysis of
the reaction of 18 with diethylamine showed that two
products 19 and 20 were formed (Scheme 7). The experi-
mentally determined ratio of 19:20 was 63:37 based on
1H NMR analysis of the reaction mixture. X-ray crystal-
lographic analysis of 18 indicated that the carbon-
bromine bond adjacent to the N-oxide functionality is the
longer of the two (1.976 Å vs 1.952 Å, Scheme 7)
consistent with preferential nucleophilic attack adjacent
to the N-oxide functionality (and hence preferential
formation of 19).18 Analogous results were obtained from
the reaction of 18 with n-butylamine.18
Conclusion
This study was designed to clarify the mechanism of
reaction of 1 with primary amines, reflecting our interest
in determining its biological mode of action and develop-
1
ing chemical proteomic tools based on 1. The H NMR
studies with n-butylamine (a mimic of a nitrogen-bound
protein nucleophile) provide the first experimental evi-
dence in support of a reaction pathway for 1 that involves
2 equiv of the amine (Scheme 1, path A). These studies
raise the intriguing possibility that two nucleophilic
residues in close proximity18 are required for irreversible
modification of a protein by 1.22 Further biochemical
studies to test this hypothesis are ongoing in our labora-
tory. In addition, computational techniques coupled with
X-ray crystallographic studies have provided a rational-
ization for the differential reactivity of 1, 7, and 18 with
amines (Schemes 3, 5, and 7). They provide a clear
explanation of the influence of the N-oxide functional
group.
FIGURE 5. (HF/6-31G(d)), PCM Solvation model for CHCl3
of (a) compound 8‡ and (b) compound 2a‡. In each case, part i
illustrates the structure of the transition state for a cyclization
reaction, part ii shows the electrostatic potential surface of
the transition state (both (a) and (b) are plotted in the same
scale), and part iii provides a detailed view of the transition
state geometry. Values of the parameters are provided in the
table in Figure 5.
C(2)-C(3) in 2a‡ (-13.1°, 32.0°) and 8‡ (-4.1°, 23.8°) also
support the view that the envelope conformation adopted
by the forming five-membered ring is more distorted in
2a‡ than in 8‡. Importantly, these calculations support
the experimentally observed reactivity of 2a and 8.
Invoking an unfavorable electronic interaction between
the N-oxide oxygen atom and the developing negative
charge on the bromine atom in 2a‡ provides a concise
explanation for the observed reaction path (Scheme 5).
A steric clash between the N-oxide oxygen atom and the
bromine atom in 2a‡ presumably also contributes to the
experimental outcome.
An interesting prediction arises from this mechanistic
explanation. Reaction of the unsymmetrical 2,3-bis-
(bromomethyl)quinoxaline 1-oxide (18) with diethylamine
would be expected to proceed as follows: (i) initial
displacement of the bromine atom adjacent to the N-oxide
Experimental Section
All commercially available substrates, reagents, and sol-
vents were used without further purification unless otherwise
1
stated. H NMR spectra were recorded at 300 and 500 MHz.
13C NMR spectra were recorded at either 75 or 125 MHz.
Coupling constants (J) are given in Hz. Low- and high-
resolution mass spectral analyses were recorded in either EI,
CI, or ES operating in positive ion mode. Melting points are
uncorrected.
(22) Several examples of small molecules that covalently modify two
residues within the same protein are known. Hartman, F. C.; Wold,
F. J. Am. Chem. Soc. 1966, 88, 3890-3891. Navia, M. A.; Springer, J.
P.; Lin, T.-Y.; Williams, H. R.; Firestone, R. A.; Pisano, J. M.; Doherty,
J. B.; Finke, P. E.; Hoogsteen, K. Nature 1987, 327, 79-82. Bode, W.;
Turk, D.; Karshikov, A. Protein Sci. 1992, 1, 426-471.
(23) Hahn, W. E.; Lesiak, J. Z. Soc. Sci. Lodz., Acta Chim. 1972,
17, 201-205.
J. Org. Chem, Vol. 70, No. 13, 2005 5059

Pearson et al.
2,3-Bis(bromomethyl)quinoxaline 1,4-Dioxide, 1.5 Puri-
fied mCPBA (13.8 g, 80.0 mmol) was added to a solution of
compound 7 (5.06 g, 16.0 mmol) in dry DCM (160 mL) with
stirring at room temperature. After 42 h the reaction mixture
was diluted with DCM (1.00 L) and washed with 10% Na2CO3
solution (2 × 250 mL). The organic phase was dried (MgSO4)
and concentrated in vacuo to give a yellow solid that was
purified by flash column chromatography on silica gel (ethyl
acetate:petroleum ether, 1:4) to yield a bright yellow crystalline
solid (3.64 g, 10.5 mmol, 65%). Mp 178.5-179.0 °C (recrystal-
lized from ethyl acetate) (lit.5 mp 188.0-189.0 °C, dioxane);
IR (NaCl, Nujol) 1341, 1036, 774, and 643 cm-1; 1H NMR (300
MHz, CDCl3) δ 8.65 (m, AA′ part of the AA′XX′ system, 2H),
ether). IR (NaCl, Nujol) 774 and 696 cm-1; 1H NMR (300 MHz,
CDCl3) δ 8.07 (m, AA′ part of the AA′XX′ system, 2H), 7.80
(m, XX′ part of the AA′XX′ system, 2H), 4.93 (s, 4H); 13C NMR
(75 MHz, CDCl3) δ 150.8, 141.5, 130.9, 129.0, 30.5; MS-EI+
318 ([M]+•, 2 × 81Br, 20%), 316 ([M]+•
,
79Br + 81Br, 40), 314
([M]+•, 2 × 79Br, 20), 237 ([M - Br]+•
,
81Br, 100), 235 ([M -
Br]+•, 79Br, 100), 156 ([M - 2Br]+•, 65); HRMS-EI+ (m/z) [M]+•
calcd for C10H8N2Br2 315.9034, found 315.9035. Anal. Calcd
for C10H8N2Br2: C, 38.01; H, 2.55; N, 8.87. Found: C, 37.87;
H, 2.27; N, 8.72.
2-Butyl-2,3-dihydro-1H-pyrrolo[3,4-b]quinoxaline, 9,7
and 7,16-Di-n-butyl[1,6]diazecino[3,4-b;8,9-b′]diquinoxa-
line, 10. Prepared according to the general kinetic NMR
procedure, using compound 7 and n-butylamine.18 Also, on a
preparative scale: n-Butylamine (0.312 g, 4.26 mmol) was
added to a solution of 7 (0.449 g, 1.42 mmol) in chloroform
(40.0 mL) at room temperature under nitrogen. After 20 h the
reaction mixture was concentrated in vacuo and purified by
flash column chromatography on silica gel (ethyl acetate:
petroleum ether, 1:3) to yield 9 as a pink solid (0.200 g, 0.881
mmol, 62%), 10 as a yellow oil (0. 0260 g, 0.0601 mmol, 8%),
and 11 (trace amounts) as an unstable red oil.
7.89 (m, XX′ part of the AA′XX′ system, 2H), 4.94 (s, 4H); 13
C
NMR (75 MHz, CDCl3) δ 139.8, 137.6, 132.5, 120.7, 20.5; MS-
ES+ 373 ([M + Na]+, 2 × 81Br, 18%), 371 ([M + Na]+, 79Br +
81Br, 100), 369 ([M + Na]+, 2 × 79Br, 20); HRMS-EI+ (m/z)
[M]+• calcd for C10H8Br2N2O2 345.8952, found 345.8949. Anal.
Calcd for C10H8Br2N2O2: C, 34.51; H, 2.32; N, 8.05. Found:
C, 34.25; H, 1.95; N, 7.70. In addition compound 18 was
isolated as a white crystalline solid (0.930 g, 2.8 mmol, 18%).
(3-(Bromomethyl)quinoxalin-2-ylmethyl)butylamine
1,4-Dioxide, Intermediate 2a, and Butyl(3-(butylami-
nomethyl)quinoxalin-2-ylmethyl)amine 1,4-Dioxide, In-
termediate 3a. Prepared according to the general kinetic
NMR procedure, using compound 1 with n-butylamine.18 After
522 s formation of intermediate 2a was observed in the
presence of n-butylamine.18 Attempts to isolate 2a proved
unsuccessful. Intermediate 2a: 1H NMR (500 MHz, CDCl3) δ
8.67-8.60 (m, 2H), 7.90-7.82 (m, 2H), 4.99 (s, 2H), 4.26 (s,
2H), 2.77-2.67 (m, 2H), 1.54-1.48 (m, 2H), 1.39-1.33 (m, 2H),
0.92-0.89 (m, 3H). After 54 572 s intermediate 3a was
observed in the presence of 5a and excess n-butylamine.18
Attempts to isolate 3a proved unsuccessful. Intermediate 3a:
1H NMR (500 MHz, CDCl3) δ 8.62 (m, AA′ part of the AA′XX′
system, 2H), 7.84 (m, XX′ part of the AA′XX′ system, 2H), 4.28
(s, 4H), 2.70 (t, J ) 7.2 Hz, 4H), 1.54-1.34 (m, 8H), 0.91 (t, J
) 7.4 Hz, 6H).
Compound 9: mp 54.5 °C dec; 1H NMR (300 MHz, CDCl3) δ
8.07-8.01 (m, AA′ part of the AA′XX′ system, 2H), 7.75-7.69
(m, XX′ part of the AA′XX′ system, 2H), 4.13 (s, 4H), 2.85 (t,
J ) 7.5 Hz, 2H), 1.70-1.60 (m, 2H), 1.52-1.39 (m, 2H), 0.99
(t, J ) 7.3 Hz, 3H); 13C NMR (75 MHz, CDCl3) δ 156.9, 141.7,
129.1, 128.9, 57.9, 56.3, 30.3, 20.5, 14.0; MS-APCI+ 270 ([M
+ 1 + CH3CN]+, 20%), 269 ([M + CH3CN]+, 100), 229 ([M +
1]+, 15), 228 ([M]+, 90); HRMS-CI+ (m/z) [M]+ calcd for
C14H18N3 228.1501, found 228.1497.
Compound 10: 1H NMR (300 MHz, CDCl3) δ 8.05-8.00 (m,
AA′ part of the AA′XX′ system, 4H), 7.72-7.66 (m, XX′ part of
the AA′XX′ system, 4H), 4.37 (s, 8H), 2.60 (t, J ) 7.7 Hz, 4H),
1.37-1.27 (m, 4H), 1.07-0.95 (m, 4H), 0.64 (t, J ) 7.3 Hz, 6H);
13C NMR (75 MHz, CDCl3) δ 154.5, 140.7, 129.3, 128.5, 63.1,
54.4, 28.9, 20.5, 13.7; MS-APCI+ 456 ([M + 1]+, 28%), 455
([M]+, 100); HRMS-CI+ (m/z) [M]+ calcd for C28H35N6 455.2923,
found 455.2915.
2-Butyl-2H-pyrrolo[3,4-b]quinoxaline, 11. Crude isola-
tion afforded a red oil: 1H NMR (500 MHz, CDCl3) δ 7.99-
7.96 (m, AA′ part of the AA′XX′ system, 2H), 7.61 (s, 2H), 7.55-
7.52 (m, XX′ part of the AA′XX′ system, 2H), 4.47 (t, J ) 7.2
Hz, 2H), 2.03 (m, 2H), 1.65-1.20 (contains m, 2H), 0.97 (t, J
) 7.4 Hz, 3H); MS-EI+ 225 ([M]+•, 33%), 183 (100), 156 (26);
HRMS-EI+ (m/z) [M]+• calcd for C14H15N3 225.1266, found
225.1276.
2-Butyl-2H-pyrrolo[3,4-b]quinoxaline 4-oxide, 5a. Pre-
pared according to the general kinetic NMR procedure, using
compound 1 and n-butylamine.18 Also, on preparative scale:
A solution of n-butylamine (0.027 g, 0.369 mmol) in deuterated
chloroform (3.00 mL) was added to a solution of 1 (0.040 g,
0.121 mmol) in deuterated chloroform (2.86 mL) with stirring
at room temperature. After 20 h the reaction mixture was
quenched with silica and the solvent removed in vacuo.
Purification by flash column chromatography on silica gel
(ethyl acetate:petroleum ether, 1:9 to 1:4) yielded an unstable
dark red oil (0.025 g, 0.10 mmol, 90%). 5a was observed to
decompose rapidly in the absence of solvent. IR (NaCl) 1578,
2,2-Diethyl-2,3-dihydro-1H-pyrrolo[3,4-b]quinoxalin-2-
ium Bromide, 14.23 Prepared according to the general kinetic
NMR procedure, using compound 7 and diethylamine.18 Reac-
tion was scaled up 30-fold for full analysis. 14 was collected
following crystallization from chloroform to yield a white
crystalline solid (0.068 g, 0.32 mmoL, 74%), mp 195.5-196.0
°C dec (lit.23 195.0 °C); IR (KBr) 3461, 3394, 2974, 2936, 1503,
1
1544, 1326, and 745 cm-1; H NMR (500 MHz, CDCl3) δ 8.55
(d, J ) 8.5 Hz, 1H), 8.00 (d, J ) 8.8 Hz, 1H), 7.67 (d, J ) 2.3
Hz, 1H), 7.61 (d, J ) 2.3 Hz, 1H), 7.59-7.56 (m, 1H), 7.53-
7.50 (m, 1H), 4.39 (t, J ) 7.1 Hz, 2H), 2.02-1.98 (m, 2H), 1.46-
1.34 (m, 2H), 0.97 (t, J ) 7.4 Hz, 3H); 13C NMR (75 MHz,
CDCl3) δ 145.0, 137.9, 133.5, 130.0, 129.0, 127.9, 125.2, 118.3,
111.4, 103.6, 53.0, 33.4, 19.7, 13.4; MS-CI+ 242 ([M + 1],
100%); HRMS-EI+ (m/z) [M]+• calcd for C14H15N3O 241.1215,
found 241.1215.
1
1465, 1030, and 763 cm-1; H NMR (300 MHz, D2O) δ 8.10-
8.04 (m, AA′ part of the AA′XX′ system, 2H), 7.93-7.87 (m,
XX′ part of the AA′XX′ system, 2H), 3.80 (q, J ) 7.2 Hz, 4H),
1
1.42 (t, J ) 7.2 Hz, 6H); H NMR (300 MHz, CDCl3) δ 8.14-
8.08 (m, AA′ part of the AA′XX′ system, 2H), 7.84-7.79 (m,
XX′ part of the AA′XX′ system, 2H), 5.59 (s, 4H), 4.15 (q, J )
7.2 Hz, 4H), 1.44 (t, J ) 7.2 Hz, 6H); 13C NMR (75 MHz, D2O,
referenced to dioxane) δ 149.6, 142.2, 132.4, 129.0, (65.2, 64.9,
64.6, 64.3, 64.0),18 59.3, 8.8; MS-ES+ 228 ([M]+, 100%), 229
([M + 1]+ 14); HRMS-CI+ (m/z) [M]+ calcd for C14H18N3
228.1501, found 228.1497.
2,3-Bis(bromomethyl)quinoxaline, 7.13 A solution of 1,2-
phenylenediamine (6.81 g, 63.0 mmol) in dry THF (40.0 mL)
was added to 1,4-dibromo-2,3-butanedione (14.6 g, 60.0 mmol)
in dry THF (80.0 mL) at 0 °C over 15 min with stirring. The
reaction was warmed to room temperature and stirred for a
further 17 h. After concentration in vacuo, the crude material
was partitioned between 10% NaHCO3 solution (150 mL) and
DCM (200 mL). The organic phase was washed with brine (1
× 100 mL), dried (MgSO4), and concentrated in vacuo to give
a dark brown solid that was purified by flash column chro-
matography on silica gel (ethyl acetate:petroleum ether, 1:9)
to yield a white crystalline solid (16.7 g, 52.9 mmol, 88%), mp
153.0 °C (sharp, recrystallized from ethyl acetate:petroleum
7,16-Di-n-butyl[1,6]diazecino[3,4-b;8,9-b′]diquinoxa-
line 5,9,14,18-Tetraoxide, 15. Prepared according to the
general kinetic NMR procedure, using compound 1 and n-
butylamine.18 Attempts to isolate 15 by column chromatogra-
phy yielded an unstable red oil (0.0015 g, 0.003 mmol, 5%).
1H NMR (500 MHz, CDCl3) δ 8.67-8.53 (m, 4H), 7.90-7.71
(m, 4H), 4.71 (s, 8H), 2.77-2.67 (m, 4H), 1.54-1.48 (m, 4H),
5060 J. Org. Chem., Vol. 70, No. 13, 2005

Controlling the Outcome of an N-Alkylation Reaction
1.39-1.33 (m, 4H), 0.92-0.89 (m, 6H); MS-ES+ 542 ([M + 1
+ Na]+, 10%), 541 ([M + Na]+, 100), 519 ([M + 1]+, 35); HRMS-
ES+ (m/z) [M + 1]+ calcd for C28H35N6O4 519.2720, found
519.2711.
fold for full analysis. Compound 19 was collected following
crystallization to yield a white crystalline solid (0.098 g, 0.30
mmol, 50%). Compound 20 was recovered from the filtrate in
the presence of diethylamine.
(3-(Diethylaminomethyl)quinoxalin-2-ylmethyl)dieth-
ylamine 1,4-Dioxide, 17. Prepared according to the general
kinetic NMR procedure, using compound 1 and diethylamine.18
Attempts to isolate 17 on a preparative scale were unsuccessful
due to rapid decomposition. 1H NMR (300 MHz, CDCl3) δ
8.68-8.62 (m, AA′ part of the AA′XX′ system, 2H), 7.84-7.78
(m, XX′ part of the AA′XX′ system, 2H), 4.35 (s, 4H), 2.64 (q,
J ) 7.1 Hz, 8H), 1.03 (t, J ) 7.1 Hz, 12H); 13C NMR (75 MHz,
CDCl3) δ 144.1, 137.2, 131.4, 120.7, 47.8, 42.7, 12.1; MS-ES+
355 ([M + Na]+, 20%), 260 ([M - N(CH2CH3)2]+, 100), 214 (14),
198 (84); HRMS-ES+ (m/z) [M + 1]+ calcd for C18H29N4O2
333.2291, found 333.2294.
Compound 19: mp 203.0-204.0 °C dec; IR (KBr) 3490, 3423,
1
2984, 2917, 1589, 1565, 1498, 1369, 1097, and 772 cm-1; H
NMR (500 MHz, D2O) δ 8.46 (dd, J ) 8.7, 1.3 Hz, 1H), 8.16
(dd, J ) 8.4, 1.3 Hz, 1H), 8.01-7.92 (m, 2H), 3.81 (q, J ) 7.2
Hz, 4H), 1.42 (t, J ) 7.2 Hz, 6H); 1H NMR (500 MHz, CDCl3)
δ 8.53 (m, 1H), 8.14 (m, 1H), 7.91-7.80 (m, 2H), 5.65 (s, 2H),
5.56 (s, 2H), 4.22-4.11 (m, 4H) 1.50 (t, J ) 7.1 Hz, 6H); 13C
NMR (125 MHz, D2O, reference to dioxane) δ 151.6 (2 × C),
146.3, 136.7, 136.3, 133.9, 132.7, 129.9, 118.2, ((66.0, 65.8, 65.7,
65.5, 65.3), (62.4, 62.2, 62.0, 61.8, 61.5)),18 59.8, 8.8; MS-ES+
244 ([M]+, 100%), 198 (81); HRMS-ES+ (m/z) [M]+ calcd for
C14H18N3O 244.1450, found 244.1456.
2,3-Bis(bromomethyl)quinoxaline 1-Oxide, 18.5 Purified
mCPBA (0.569 g, 3.30 mmol) was added to a solution of 7 (1.01
g, 3.20 mmol) in dry DCM (20.0 mL) at room temperature.
After 20 h the reaction mixture was washed with 10% Na2-
CO3 (2 × 20 mL) and the combined aqueous washings
extracted with DCM (3 × 50 mL). The combined organic phase
was dried (MgSO4) and concentrated in vacuo to give a pale
yellow solid that was purified by flash column chromatography
on silica gel (ethyl acetate:petroleum ether, 1:9 to 1:4) to yield
a white crystalline solid (0.548 g, 1.65 mmol, 52%). Mp 168.5-
169.0 °C (recrystallized from ethyl acetate:petroleum ether)
(lit.5 167.0-168.0 °C, ethanol); IR (NaCl, Nujol) 1564, 1481,
1357, 1060, 769, 675, and 662 cm-1; 1H NMR (300 MHz, CDCl3)
δ 8.57 (dd, J ) 8.5, 1.5 Hz, 1H), 8.54 (dd, J ) 8.4, 1.5 Hz, 1H),
7.87-7.73 (m, 2H), 5.02 (s, 2H), 4.77 (s, 2H); 13C NMR (75
MHz, CDCl3) δ 152.6, 143.4, 138.5, 136.3, 132.2, 130.9, 130.1,
119.2, 30.2, 20.8; MS-ES+ 357 ([M + Na]+, 2 × 81Br, 19%),
355 ([M + Na]+, 79Br + 81Br, 100), 353 ([M + Na]+, 2 × 79Br,
23); HRMS-EI+ (m/z) [M]+• calcd for C10H8Br2N2O 329.9003,
found 329.9002. Anal. Calcd for C10H8Br2N2O: C, 36.18; H,
2.43; N, 8.44. Found: C, 36.18; H, 2.07; N, 8.14. Compounds
1 (0.052 g, 0.15 mmol, 5%) and 7 (0.411 g, 1.3 mmol, 41%)
were also isolated.
1
Compound 20: H NMR (300 MHz, CDCl3) δ 8.58 (m, 1H),
8.08 (m, 1H), 7.78-7.65 (m, 2H), 4.34 (s, 2H), 4.13 (s, 2H),
2.69-2.59 (m, 8H), 1.06-1.00 (m, 12H); MS-ES+ 339 ([M +
Na]+, 32%), 317 ([M + 1]+, 17), 299 ([M - OH]+, 24), 244 ([M]+,
compound 19, 94), 228 ([M - O]+, compound 19, 100); HRMS-
ES+ (m/z) [M + 1]+ calcd for C18H29N4O 317.2341, found
317.2354.
Acknowledgment. The authors thank the NIH (No.
R01A1054961 (R.J.P.) PI Professor Gary Ward (Uni-
versity of Vermont), co-PI (N.J.W.)) and BBSRC (No.
02/A1/B/08395 (K.M.E.)) for funding, Mrs M. Smith and
Dr. T. Lebl for NMR assistance, Mrs C. Horsburgh for
HRMS analysis, and Mrs S. Williamson for elemental
analysis. We thank Professor Gary Ward for useful
discussions regarding the biological activity. We are
grateful to Dr David Smith for critical appraisal of this
manuscript and Professor David O’Hagan for sugges-
tions made during the course of this research.
Supporting Information Available: 1H and 13C charac-
terization, additional schemes, kinetic profiles, Cartesian
coordinates, and total energies for the optimized structures of
the calculated transition states. This material is available free
of charge via the Internet at http://pubs.acs.org.
2,2-Diethyl-2,3-dihydro-1H-pyrrolo[3,4-b]quinoxalin-2-
ium Bromide, 4-Oxide, 19, and (3-(Diethylaminomethyl)-
quinoxalin-2-ylmethyl)diethylamine 4-Oxide, 20. Pre-
pared according to the general kinetic NMR procedure, using
compound 18 and diethylamine.18 Reaction was scaled up 30-
JO0503106
J. Org. Chem, Vol. 70, No. 13, 2005 5061