Ambident reactivity of aryloxide ions towards the super-electrophile, 4,6-dinitrobenzofuroxan, Kinetics, thermodynamics and stereoelectronic factors on regioselectivity

Article abstract of DOI:10.1039/a605907d

Reactions of the aryloxide ions, phenoxide (PhO^-) and 3,5-di-fert-butylphenoxide (3,5-DTBPhO^-), with the super-electrophilic heteroaromatic substrate, 4,6-dinitrobenzofuroxan (DNBF, 1), have been examined by 400 MHz ¹H NMR spectroscopy in acetronitrile-dimethoxyethane ([²H₃]MeCN:[²H₁₀]DME 1:1, v/v) as a function of varying temperature (-40 to 23°C) and in dimethyl sulfoxide ([²H₆]DMSO) at room temperature. We herein report the first observation and full characterization of the O-bonded σ-adduct (DNBF·OPh^-, 3a) formed by attack of PhO^-, acting as an O-nucleophile, at the C-7 super-electrophilic site of 1. No C-5 adduct was seen in the initial spectrum (- 40°C, [²H₃]MeCN:[²H₁₀]DME) or in subsequent monitoring of the reaction. These results suggest that PhO^- displays K7T7 regioselectivity towards DNBF wherein attack at the C-7 site is favoured by both kinetics and thermodynamics, comparable to the behaviour shown by PhO^- towards 2,4,6-trinitroanisole where the C-1 adduct is the product of both kinetic and thermodynamic control (i.e. K1T1 regioselectivity). Upon warming the reaction mixture to ambient, the C-7 O-adduct, DNBF·OPh^-, 3a, gives way to the more stable C-7 C-bonded σ-adducts (DNBF-ortho-PhOH^- adduct, 4, and DNBF·para-PhOH^- adduct, 5, in a ratio of ca. 1:6). The C-7 hydroxide adduct, DNBF-OH^- , 2a, and phenol are detected at this temperature. The C-adducts, 4 and 5, are the sole PhO^- adducts previously observed in the DNBF-PhO^- reaction system (in [²H₆]DMSO at room temperature). When C-attachment is precluded by steric hindrance, as in the reaction of 1 with 3,5-DTBPhO^-, the C-7 DNBF·OPhDTB^- adduct, 3b, is observed by ¹H NMR spectroscopy even in [²H₆]DMSO under ambient conditions. The results of the kinetics and thermodynamics of aryloxide adduct formation with DNBF, including the ambident reactivity found, are discussed with regard to stereoelectronic stabilization in the adducts and with comparison to relevant 4-nitrobenzofuroxan (NBF), 2-(nitroaryl)-4,6-dinitrobenzotriazole 1-oxides (2-Ar-4,6-DNBT) systems and to the normal electrophile, 1,3,5-trinitrobenzene (TNB).

Full text of DOI:10.1039/a605907d

View Article Online / Journal Homepage / Table of Contents for this issue
Ambident reactivity of aryloxide ions towards the super-electrophile,
4,6-dinitrobenzofuroxan. Kinetics, thermodynamics and
stereoelectronic factors on regioselectivity†
Erwin Buncel,a Richard A. Mandervilleb and Julian M. Dustc
a
Department of Chemistry, Queen’s University, Kingston, Ontario, Canada K7L 3N6
b Department of Chemistry, Wake Forest University, Winston-Salem, North Carolina,
USA 27109
c Department of Chemistry, Sir Wilfred Grenfell College, Corner Brook, Newfoundland,
Canada A2H 6P9
Reactions of the aryloxide ions, phenoxide (PhOϪ) and 3,5-di-tert-butylphenoxide (3,5-D TBPhOϪ), with
the super-electrophilic heteroaromatic substrate, 4,6-dinitrobenzofuroxan (D N BF, 1), have been examined
by 400 M H z 1H N M R spectroscopy in acetronitrile–dimethoxyethane ([2H 3]M eCN :[2H 10]D M E 1 :1, v/v)
as a function of varying temperature (Ϫ40 to 23 ЊC) and in dimethyl sulfoxide ([2H 6]D M SO) at room
temperature. We herein report the first observation and full characterization of the O-bonded σ-adduct
(D N BF ؒOPhϪ, 3a) formed by attack of PhOϪ, acting as an O-nucleophile, at the C-7 super-electrophilic site
of 1. N o C-5 adduct was seen in the initial spectrum (Ϫ40 ЊC, [2H 3]M eCN :[2H 10]D M E) or in subsequent
monitoring of the reaction. These results suggest that PhOϪ displays K 7T7 regioselectivity towards
D N BF wherein attack at the C-7 site is favoured by both kinetics and thermodynamics, comparable to the
behaviour shown by PhOϪ towards 2,4,6-trinitroanisole where the C-1 adduct is the product of both
kinetic and thermodynamic control (i.e. K 1T1 regioselectivity). U pon warming the reaction mixture to
ambient, the C-7 O-adduct, D N BF ؒOPhϪ, 3a, gives way to the more stable C-7 C-bonded σ-adducts
(D N BF ؒortho-PhOH Ϫ adduct, 4, and D N BF ؒpara-PhOH Ϫ adduct, 5, in a ratio of ca. 1 :6). The C-7
hydroxide adduct, D N BF ؒOH Ϫ, 2a, and phenol are detected at this temperature. The C-adducts, 4 and 5,
are the sole PhOϪ adducts previously observed in the D N BF –PhOϪ reaction system (in [2H 6]D M SO at
room temperature). When C-attachment is precluded by steric hindrance, as in the reaction of 1 with
3,5-D TBPhOϪ, the C-7 D N BF ؒOPhD TBϪ adduct, 3b, is observed by 1H N M R spectroscopy even in
[2H 6]D M SO under ambient conditions. The results of the kinetics and thermodynamics of aryloxide
adduct formation with D N BF, including the ambident reactivity found, are discussed with regard to
stereoelectronic stabilization in the adducts and with comparison to relevant 4-nitrobenzofuroxan (N BF ),
2-(nitroaryl)-4,6-dinitrobenzotriazole 1-oxides (2-Ar-4,6-D N BT) systems and to the normal electrophile,
1,3,5-trinitrobenzene (TN B).
1
TNB,10a using H NMR spectroscopy in the novel medium,
Introduction
acetonitrile–dimethoxyethane ([2H3]MeCN :[2H10]DME 1:1
v/v); this solvent system remains fluid to temperatures of
Ϫ50 ЊC and, so, facilitated observation of these reactions at
Ϫ40 ЊC. As a comparison of the reactivity of aryloxide ions
towards NBF and TNB, we now report our findings on the
reactivity of PhOϪ and 3,5-di-tert-butylphenoxide (3,5-
DTBPhOϪ) with the super-electrophilic heteroaromatic, 4,6-
dinitrobenzofuroxan (DNBF, 1).
Current interest in the reactivity of 1 has been fuelled by the
fact that this neutral 10 π-electron heteroaromatic substrate is a
more powerful electrophile than such strong electrophiles as the
p-nitrobenzenediazonium cation and the proton.11 Thus, water
and methanol react readily with 1 [eqn. (1)] to yield the C-7
Studies of the interactions of nucleophilic reagents with poly-
nitroaromatics and heteroaromatics have led to the discovery of
a diverse number of species (π-complexes, and more notably
radical anions,1,2g radical anion–radical pairs1b,c and anionic
σ-adducts2) that have yielded valuable information concern-
ing reaction mechanisms. Electron-deficient nitroaromatic and
heteroaromatic compounds have been used for the identifi-
cation of thiols3 and amino (usually lysine) nucleophilic resi-
dues4 in proteins. The anionic σ-adducts that have frequently
been postulated as intermediates in these reactions have
organic synthetic utility in their own right.2a The nitroaromatic
and heteroaromatic substrates also serve as probes of ambident
nucleophilic reactivity in anionic σ-adduct formation,2 and, in
this regard, our 5–7 and other groups8 have been particularly
interested in the reactivity of phenoxide ion (PhOϪ) as an
oxygen- and carbon-nucleophile towards 1,3,5-trinitrobenzene
(TNB),5,8 2,4,6-trinitroanisole (TNA),2a,6 and the 2-(nitroaryl)-
4,6-dinitrobenzotriazole 1-oxide (2-Ar-4,6-DNBT)7 series of
ambident electrophiles. Recently, we have fully characterized
the O-bonded anionic σ-adducts formed by PhOϪ with 4-nitro-
benzofuroxan (NBF)9 and the normal reference electrophile,
O–
O–
H
OR
O2N
O2N
H
N
+
N
+
O
+
+
H+ (1)
ROH
O
N
N
–NO2
NO2
1
R = H
R = CH3
2a
2b
hydroxide and methoxide adducts, 2a and 2b, respectively,
which are 1010 times more stable than the analogous TNBؒOHϪ
and TNBؒOMeϪ adducts that are only formed upon addition of
hydroxide and methoxide ions.1,2
† For Part 52 of the series on anionic σ-complexes, see ref. 2(h). For
Part 11 of the series on heteroaromatic super-electrophiles in σ-adduct
formation, see ref. 24.
J. Chem. Soc., Perkin Trans. 2, 1997
1019

View Article Online
recorded during our previous investigation of the DNBF–PhOϪ
reaction system in [2H6]DMSO,7b the resonances of the domin-
ant σ-complex (i.e. those found at δ 5.40 and 8.64) could be
attributed to the para C-bonded DNBFؒPhOHϪ adduct, 5
(Scheme 1). Equally important, the phenoxy ring protons of 5
O–
O–
O2N
N
+
O
+
N
R
R
NO2
R = H, PhO
R = But, 3,5-DTBPhO
1
Fig. 1 1H NMR spectrum (downfield region; 5.4–9.6 ppm) of the
DNBF–PhOK (1:1) reaction, recorded at ca. 3 min reaction time
[2H6]DMSO at room temperature). The resonances that are highlighted
are assigned to the C-7 adducts (para and ortho C-adducts and the
hydroxide adduct). The singlet marked DBB represents 1,4-
dibromobenzene, an internal integration standard.
HO
The ease of σ-adduct formation with 1 has enabled its use to
assess the reactivity of weakly nucleophilic centres.7b,12–14 In this
context, 1 reacts according to a formal SEAr mechanism 15 with
the weakly nucleophilic benzene ring of 1,3,5-trimethoxy-
benzene;14 similar facile substitution reactions proceed with
π-electron rich heteroaromatics including indoles,11 pyrrole,
thiophene and furan.16
R
R
OH
O–
O–
H
O
H
O2N
O2N
N
N
+
+
O
O
N
N
In the present paper, we describe the reaction of PhOϪ and
3,5-DTBPhOϪ towards DNBF. For both anions we have been
able to thoroughly characterize the oxygen-bonded anionic
σ-adducts (i.e. 3a and 3b in Scheme 1) using 400 MHz 1H NMR
spectroscopy. The spectroscopic characteristics of the C-7
DNBFؒOPhϪ adduct, 3a, were acquired in MeCN–DME at
Ϫ40 ЊC whereas the features of the C-7 DNBFؒOPhDTBϪ
adduct, 3b, could be obtained in DMSO at ambient tem-
perature. These results confirm our surmise that formation of
O-adducts of PhOϪ with 1 is kinetically favoured over form-
ation of the corresponding para and ortho C-bonded adducts.7b
Further, the regioselectivity shown by PhOϪ, as an O-nucleo-
phile, will be discussed in terms of patterns of regioselectivity
previously classified for picryl ether systems.6,17 The regio-
selectivity found depends partly on the stereoelectronic stabil-
ization available to the relevant adducts and stereoelectronic
factors affecting the kinetics and thermodynamics of adduct
formation will be discussed. Finally, the enhanced stability of
the DNBF O-adducts, as compared to corresponding O-
adducts of NBF and TNB, reemphasizes the super-electro-
philicity of 1 and allows us to draw comparisons between these
systems and the 2-nitroaryl-4,6-dinitrobenzotriazole 1-oxide
series that possesses both normal and super-electrophilic reac-
tion sites.
–NO2
–NO2
4
3a
R = H
R = But 3b
R = H
HO
O–
H
O2N
N
+
O
N
–NO2
5
R = H
Scheme 1
appear in the δ 6.74–7.16 region, while the phenolic OH reson-
ates at δ 9.54. The identity of the minor product whose signals
are located at δ 5.62 and 8.66 is consistent with formation of the
ortho C-bonded adduct, 4,7b whereas peaks at δ 5.84 and 8.59
stem from production of the DNBFؒOHϪ adduct, 2a.18
The presence of adventitious water in the [2H6]DMSO sol-
vent, coupled with general base catalysis by phenoxide ion,
accounts for the formation of 2a and phenol (PhOH). Note that
the H-7 signals ascribed to the C-adducts (δ 5.62, 5.40) appear
upfield of the H-7 signal assigned to the DNBFؒOHϪ adduct (δ
5.84). This is in accord with the effect of the greater electro-
negativity of the oxygen centre in 2a on the chemical shift of
H-7 of this adduct as compared to the electronegativity of the
attached C-centres in 4 and 5. The assignment of the C-7 sig-
nals of the carbon adducts is also consistent with the chemical
shift data recently reported for a series of carbon adducts of
DNBF formed by nitrocarbanion attack at C-7.10c
Results
Reaction of 1 with 1 equiv. of phenoxide ion in DMSO
Addition of 1 equiv. of PhOϪ (as potassium phenoxide solu-
tion, PhOK) in [2H6]DMSO to a [2H6]DMSO solution of
DNBF 1 (final concentration: 0.09 ), immediately produced a
1
deep-red solution. Fig. 1 shows a H NMR spectrum of the
reaction mixture acquired ca. 3 min after the reagents were
mixed. Apparent in this first spectrum are resonances that
represent three distinct σ-adducts. The H-7 proton in a given
DNBF moiety is bound to an sp3-hybridized carbon centre; the
chemical shift of this signal is notably sensitive to the nature of
the group attached to this position.2,7,10 It is, therefore, signifi-
cant that three different resonances are observed in the region
of the spectrum typical for H-7 in a DNBF σ-complex,10b,c
namely: δ 5.84 (21%), 5.62 (7%) and 5.40 (43%) with signals for
the corresponding H-5 protons being found at δ 8.59, 8.66 and
8.64, respectively. On the basis of comparison with spectra
Few changes in the spectra of the system were noted over a
2 h period. Upon acidification (5 µl trifluoroacetic acid; TFA)
peaks attributed to the C-bonded adducts, 4 and 5, remain
unchanged, while the resonances belonging to the DNBFؒOHϪ
adduct, 2a, decreased in intensity, but did not vanish. These
observations support the present assignments. Thus, anionic σ-
adducts formed by C-attack of PhOϪ on any electron-deficient
substrate are generally stable to acid, while O-adducts formed
1020
J. Chem. Soc., Perkin Trans. 2, 1997

View Article Online
selective isotopic labelling, the upfield shift of H-7 was 3.38
ppm (i.e. δ 9.27 for DNBF–H-7 Ϫδ 5.89 for H-7 of DNBFؒ
OMeϪ) whereas the upfield shift of H-5 was only 0.25 ppm (δ
8.95 for DNBF–H-5 Ϫδ 8.70 for H-5 of DNBFؒOMeϪ).10b
Based on comparison with the picryl ether systems2,17
attachment of a phenoxy group to a ring site is expected to
cause an upfield shift of ca. 2.5 ppm. Thus, attachment at C-7
should yield a signal at δ 6.77 and attachment at C-5 should
yield a signal at δ 6.45. The signal found in the DNBF–PhOϪ
system at δ 6.78 is, therefore, likely due to H-7 and the
adduct is a C-7 DNBFؒOPhϪ adduct and not a C-5 adduct.
Further support for the assignment will emerge in the
Discussion.
Further monitoring of the reaction, by allowing the temper-
ature to rise in 5 ЊC increments after ca. 30 min, revealed that
the signals attributed to 3a were dominant in the spectra
recorded from Ϫ40 ЊC up to Ϫ20 ЊC. Above this temperature
peaks that belong to the para C-bonded DNBFؒPhOϪ adduct,
5, began to increase in intensity, coupled with the appearance
of small signals identified as belonging to the ortho C-bonded
DNBFؒPhOϪ adduct and the decline in intensity of those
assigned to 3a. At 0 ЊC, peaks due to 3a had completely van-
ished in favour of those ascribed to the C-adducts 4 and 5. On
further warming to ambient temperature (ca. 23 ЊC), and at
longer times (>2 h), the only species that remained in solution
were the C-adducts, 4 and 5, and the DNBFؒOHϪ adduct, 2a,
i.e. the same final products observed in the DMSO experiment.
At no time during this sequence could signals attributable to a
C-5 phenoxide O- or C-adduct be observed.
Fig. 2 1H NMR spectrum (aromatic region; 6.8–8.8 ppm) of the
DNBF–PhOK (1:1) reaction, recorded at ca. 3 min reaction time
([2H3]MeCN–[2H10]DME at Ϫ40 ЊC). The spectrum contains the reson-
ances ascribed to the phenoxide C-7 DNBF·OPhϪ O-adduct, 3a.
from normal electrophiles such as TNB and TNA are acid-
labile.1,2,5–7,9,10a On the other hand, even O-centred adducts of
DNBF are acid stable‡ relative to the corresponding TNB com-
plexes. In this regard, the C-7 DNBFؒOMeϪ adduct does not
decompose even at a pH of 5, whereas the TNBؒOMeϪ complex
readily reverts to TNB and methanol at ca. pH 7.19 In fact, the
DNBF ring is such
a powerful electron sink that the
DNBFؒOHϪ adduct is quite acidic (pKa = 1.5 in 90% DMSO–
H2O).20
In [2H6]DMSO (ambient temperature), signals that could be
ascribed to formation of the C-7 DNBFؒOPhϪ adduct, 3a
(Scheme 1) were not observed, in agreement with our previous
investigation.7b Moreover, no C-5 O- or C-adducts of 1 could be
observed either initially or throughout the period of obser-
vation. To detect these putative species, the reaction was
repeated in [2H3]MeCN :[2H10]DME (1:1 v/v) at reduced
temperatures.
The interaction of DNBF with PhOϪ at low temperatures in
MeCN–DME as described above, has confirmed the apparent
kinetic preference for O-attachment in reactions of DNBF, 1,
with aryloxide nucleophiles. In an attempt to examine O-adduct
formation in DMSO at ambient temperature, the reaction of
1 with 3,5-di-tert-butylphenoxide ion (3,5-DTBPhOϪ) was
examined. In this system, C-attachment is sterically inaccessible
owing to the bulky tert-butyl groups that would be expected to
block attack involving the positions either ortho or para to the
aryloxyl oxygen;9 such a condition should favour detection of
any O-adducts.
Reaction of 1 with equimolar phenoxide ion in MeCN–DME
To a solution of DNBF in [2H3]MeCN–[2H10]DME (1:1, v/v)
cooled to Ϫ40 ЊC was injected a similarly cooled solution of
phenoxide ion (PhOK in the same medium; final concentrations
ca. 0.1 ). Fig. 2 reproduces the initial spectrum of the reaction
mixture as recorded at Ϫ40 ЊC (acquired within approximately
3 min of mixing). Remarkably, the spectrum was clean and the
intense resonances could be assigned to a single complex, the
C-7 O-bonded DNBFؒOPhϪ adduct, 3a, having peaks at δ 8.77
(H-5, s), 7.24 (H-meta, m), 7.10 (H-ortho, m), 6.96 (H-para, m)
and 6.78 (H-7, s). As in previous cases, the integrals supported
the assignments made. As expected from our previous studies
of electrophile–aryloxide systems,9,10a,21 the chemical shift of
the sensitive sp3-attached H-7 proton in 3a (δ 6.78) was found
downfield (ca. 0.9 ppm) with respect to its counterpart
proton in the C-7 DNBFؒOHϪ adduct, 2a, (δ 5.84)18 or the
C-7 DBNFؒOMeϪ adduct, 2b, (δ 5.87),1,10b which reflects the
electron-withdrawing nature of the phenyl group.9
Reaction of 1 with equimolar 3,5-di-tert-butylphenoxide ion in
DMSO
Injection of 1 equiv. of a [2H6]DMSO solution of 3,5-di-tert-
butylphenoxide ion (as 3,5-DTBPhOK) into an NMR tube that
contained a [2H6]DMSO solution of 1 (final concentration 0.1
) yielded a deep red mixture. Assessment of the first spectrum
revealed that signals attributable to formation of the C-7 aryl-
oxide O-adduct, 3b (Scheme 1), were present (43%). In
[2H6]DMSO (ambient temperature) the 1H NMR signals of 3b
are as follows: δ 8.66 (H-5, s), 7.01 (H-4Ј, t, J 1.5 Hz), 6.71 (H-7,
s), 6.68 (H-2Ј,6Ј, d, J 1.5 Hz) and 1.19 (But groups, s). Again,
the integral ratios supported the assignments made. The down-
field shift of the H-7 resonance at 6.71 ppm accords with
expectations derived from characterization of the C-7 DNBFؒ
OPhϪ adduct, 3a, in [2H3]MeCN–[2H10]DME and from related
systems,9,10a,20 corrected for the slight solvent-dependency of
these chemical shifts.6 Other peaks in the spectrum belong to
3,5-DTBPhOH (47%) and the C-7 DNBFؒOHϪ adduct; these
arise from equilibration of the aryloxide with adventitious H2O
in the [2H6]DMSO medium.6,7b After 30 min reaction time,
resonances ascribed to 3b were no longer present in the spec-
trum. Instead, the spectrum contained signals for 2a, 3,5-
DTBPhOH and unmodified substrate, 1.
As in the DMSO experiment, there was no evidence for the
formation of a C-5 O-centred DNBFؒOPhϪ adduct even in this
initial spectrum. However, comment need be made here about
the assignment of adducts as C-7 rather than C-5 since reaction
at the C-5 centre could give rise to a spectrum similar to that
obtained from attack at C-7 (i.e. giving an upfield signal at ca.
5–7 ppm and a downfield signal at ca. 8.5–8.8 ppm). Chemical
shift evidence, though not conclusive, is helpful in making the
assignment. The proton at the site of attachment undergoes a
major upfield shift while the proton bound to the remaining sp2
centre undergoes only a small upfield shift. In the case of the
C-7 methoxide adduct of 1, whose assignment was conclusively
In a separate experiment, the reaction mixture was acidified
(5 µl TFA) after acquisition of the initial spectrum. Subsequent
spectra showed that the signals ascribed to the C-7 DNBFؒ
OPhDTBϪ adduct, 3b, had disappeared completely, whereas
resonances belonging to the hydroxide complex, 2a, remained.
This observation demonstrates the superior thermodynamic
1
demonstrated through H, 13C and 15N NMR studies, involving
‡ A referee has noted that equilibrium protonation of 2a could involve
protonation on the C(4)–NO2 group.
J. Chem. Soc., Perkin Trans. 2, 1997
1021

View Article Online
Table 1 1H NMR spectroscopic characteristicsa of the anionic σ-
adducts of DNBF, 1, in [2H6]DMSO b and [2H3]MeCN :[2H10]DME
(1:1 v/v)c
O
OH
O
O–
N
O–
H
H
H
c
k1
Adduct H-7
H-5
Other signals (δ/ppm; J/Hz)
k2
O2N
O2N
N
O
N
c
k–1
O
2a b
3a c
5.83, s 8.59, s 6.23 (br, s, OH)
6.78, s 8.77, s 7.10 (m, 2 H, H-ortho), 7.24 (m, 2 H, H-
meta), 6.96 (m, 2 H, H-para)
N
PhO–
O
–NO2
–NO2
1
+
3bb
6.71, s 8.66, s 7.01 (t, J 1.5, 1 H, H-paraЈ), 6.68 (d, J 1.5,
2H, H-orthoЈ), 1.19(s, 18H, 3Ј,5Ј-di-But ).
5.62, s 8.66, s Obscured.d
4q
4
4b
5b
5.40, s 8.64, s 7.16, 6.74 (A2X2 d,d, partly obscured),d
9.54 (br s, OH)
5q
5
a
Chemical shifts are measured at 400.1 MHz. b Determined in [2H6]-
DMSO at ambient temperature. c Determined in [2H3]MeCN :[2H10]-
DME AT Ϫ40 ЊC. d Signals due to the ring protons of PhOϪ/PhOH, as
well as those of the ring protons of the attached phenoxyl moieties of 4
and 5 overlap, cf. Fig. 1.
Scheme 2
temperature, even though the first spectrum was acquired with-
in ca. 3 min of mixing the reagents.7b
stability of the DNBFؒOHϪ adduct, 2a, as compared to the
aryloxide O-adduct, 3b.
This 1H NMR investigation has confirmed our earlier suppo-
sitions. Even with the superior spectroscopic sensitivity avail-
able in the present study (400 versus 100 MHz instruments), no
PhOϪ O-adduct of DNBF, 1, was seen in the initial spectrum
obtained in [2H6]DMSO at room temp. (cf. Fig. 1). However, by
altering the solvent system to [2H3]MeCN :[2H10]DME (1:1 v/v)
NMR measurements could be made at reduced temperatures
(Ϫ40 ЊC).6,9,10a,22 Now, O-attachment is spectrally observable
and the C-7 DNBFؒOPhϪ O-adduct, 3a, is the only species
detected in the initial spectrum (Fig. 2). The fact that no
C-adduct was produced at Ϫ40 ЊC clearly establishes the kinetic
preference for O-attachment in reactions of ambident aryloxide
O- and C-nucleophiles with 1.
The present study has confirmed our earlier expectations6
that, under appropriate conditions, aryloxide O-adducts of a
range of electrophiles, including TNB,10 NBF 9 and, now,
DNBF, could be detected and fully characterized. The 1H
NMR chemical shifts of the σ-adducts observed in the present
study are summarized in Table 1. The nature of the ambident
reactivity of the aryloxide nucleophiles, as well as those factors
that account for observation of the C-7 O-adducts including
stereoelectronic stabilization, as compared to the C-5 regio-
isomeric adducts, will be considered below with appropriate
comparisons to related reaction systems.
It is interesting to compare these results obtained with the
super-electrophile, DNBF, to the observations gleaned from
the 2-(nitroaryl)-4,6-dinitrobenzotriazole 1-oxide series (2-Ar-
DNBT). These novel electron-deficient substrates contain both
a super-electrophilic centre, C-7, and a normal electrophilic
site, C-1Ј. Reaction of 2-(2Ј,4Ј,6Ј-trinitrophenyl)-4,6-dinitro-
benzotriazole 1-oxide (Pi–DNBT) with PhOϪ (DMSO, room
temperature) led only to displacement of the picryl moiety (i.e.
formation of phenyl 2,4,6-trinitrophenyl ether); no C-7 O- or
C-phenoxide adduct was seen, nor was a C-1Ј adduct noted
although displacement at C-1Ј would reasonably proceed
through such a species.7b–d It is pertinent to note that the C-7
centre in the 2-Ar-DNBT series has been ranked between the
comparable site in TNB and DNBF in order of increasing
electrophilicity.7 When the reactivity of the C-1Ј site was
reduced (by successive removal of nitro groups from the 2-
aryl moiety) C-7 carbon-centred adducts, para and ortho,
could now be observed, but the product of C-7 O-attack was
still not detected.7a This comparison also supports the previ-
ous arguments concerning the relative stability of phenoxide
O-adducts, even when such adducts are formed with super-
electrophiles.
Further confirmation of the formation of aryloxide
O-adducts of DNBF, as well as assistance in the spectroscopic
assignment of the phenoxide O-adduct of 1, was obtained from
examination of the DNBF–3,5-DTBPhOϪ reaction system in
DMSO (room temp.). In the 3,5-di-tert-butylphenoxide nucleo-
phile the sites ortho and para to the oxygen centre are sterically
blocked, the kinetic barrier (k1C, Scheme 2) to formation of the
thermodynamically favoured C-adduct(s) is substantially
raised, while that for O-attack is left unchanged. Now, O-
attachment of the aryloxide nucleophile can be observed in
[2H6]DMSO at ambient temperature. Significantly, the
DNBFؒOPhDTBϪ σ-adduct, 3b, was present in solution
throughout the period of observation (ca. 30 min). This con-
trasts with the results recorded in the NBF–3,5-DTBPhOϪ sys-
tem 9 and the TNB–3,5-DTBPhOϪ reaction system,10 both
examined in DMSO. In the TNB reaction, diaryl ether for-
Discussion
Oxygen- versus carbon-reactivity
In our previous NMR (100 MHz) investigation of phenoxide
reactivity towards the super-electrophile, DNBF, 1, in [2H6]-
DMSO at ambient temperature,7b evidence for O-attack was not
obtained; no DNBFؒOPhϪ adduct was observed under these
conditions. However, formation of an O-centred adduct would
be expected to be kinetically preferred over C-adduct for-
mation. Thus, O-attack by PhOϪ occurs via a single step,
whereas formation of a C-adduct plausibly proceeds in two
steps. As shown in Scheme 2 (for formation of the ortho
C-adduct) the first step (k1C) involves attachment of the phen-
oxy ring to DNBF and formation of a quinoidal adduct (4q).
The loss of aromaticity concomitant with formation of the
quinoidal C-adduct, as well as structural and solvent re-
organization, combine to indicate that this step should be
slow.5–7a,21 Both the reverse step (decomposition of the quino-
idal adduct, 4q, back to starting materials, kϪ1C) and the
forward tautomerization step (k2C) that would generate the
C-adduct (4, in this case) result in restoration of aromaticity
and on these grounds both steps would be expected to be fast.
C
However, the k2 step represents a bimolecular reaction, involv-
ing base, and would be expected to be favoured over the uni-
molecular decomposition (kϪ1C) under the alkaline conditions
of this study. In any case, further reaction of the quinoidal
adduct is likely to be rapid since the quinoidal adduct
itself is not observed in this study or related ones.5–7a,21 The
rearomatization that occurs in the second step, k2C, confers
effective irreversibility on the overall process of C-adduct
formation and establishes the C-adduct as the product of
thermodynamic control on the DNBFؒPhOϪ reaction system.
In summary, the lack of success in observing the O-bonded
adduct, 3a, was presumed to arise not from any abnormally
high kinetic barrier to its formation but could be attributed to
its transient nature. Therefore, its transformation into the ortho
and para C-adducts, 4 and 5, occurs too rapidly to permit
1
characterization by H NMR spectroscopy in DMSO at room
1022
J. Chem. Soc., Perkin Trans. 2, 1997

View Article Online
mation through NO2 group displacement, as well as competi-
tive TNBؒOHϪ adduct formation made identification and struc-
tural elucidation of the TNBؒOPhDTBϪ complex less definitive
than in the current DNBF–3,5-DTBPhOϪ study. In the reaction
of 3,5-DTBPhOϪ with NBF,9 σ-adduct formation could not be
detected in DMSO at ambient temperature as a result of the
intervention of a rapid deoxygenation pathway. This decom-
position, in which the sp3-bound proton (H-7) of the adduct
is transferred to the 1-oxide oxygen which may subsequently
be lost, has been observed previously as a competitive process
in the reaction of methoxide with NBF 23 and in the reaction
of Pi-DNBT with isopropoxide ion.24 In the case of the NBF
and TNB reactions with 3,5-DTBPhOϪ, however, the aryl-
oxide O-adducts could be seen and their structural features
assigned using the low temperature medium (MeCN–DME).
Even under these conditions the σ-adducts exhibited relatively
short lifetimes.9,10a
Clearly, the ease of observation of the DNBFؒOPhDTBϪ
O-adduct arises from a number of factors. First, formation of
the O-adduct in this system benefits from the super-electro-
philicity of DNBF 11 as compared to TNB or even NBF.
Importantly, if C-adduct formation were not blocked sterically,
then the process of C-adduct formation would also be
enhanced by using 1 relative to TNB (see below). This
super-electrophilicity has been attributed to the relatively low
aromaticity of the heteroaromatic, as well as the intrinsic
electron-withdrawing ability of the furoxan moiety as com-
pared to a standard nitro group (as in TNB).1a,2a,b Secondly,
adducts formed by DNBF are generally more stable than their
TNB counterparts. If the absolute energy of the activation
barrier for the deoxygenation step, which is observed in various
NBF and 2-Ar-DNBT systems but not in the DNBF reactions,
remains approximately constant the greater stability of the
DNBF adducts translates into a higher kinetic barrier to
decomposition by this pathway for the adducts formed by
1.
Fig. 3 Qualitative comparative energy–reaction coordinate diagrams
that describe four general patterns of regioselectivity. Kinetic barriers
and relative stabilities are exaggerated for clarity. K5T7 represents sys-
tems in which formation of the C-5 adduct is the product of kinetic
control, but the C-7 adduct is the thermodynamic product. In the K7T7
profile the C-7 adduct is favoured by both kinetics and thermo-
dynamics. In the same way, K5T5 represents the situation where the C-5
adduct is doubly preferred: by kinetics and by thermodynamics. K7T5
profile describes the inverse behaviour from that indicated by K5T7;
now, the C-7 adduct is favoured kinetically, but the C-5 adduct is the
most stable product. These profiles, derived from consideration of the
regioselectivity found in picryl ether–nucleophile systems,17 are now
extended to DNBF systems and are further described in the text.
extended to the related heteroaromatic systems. In a hypo-
thetical case, then, attack of a nucleophile at C-5 of DNBF, 1,
to form a C-5 adduct could be kinetically preferred while
attack of the same nucleophile at C-7 could be thermo-
dynamically preferred. This situation would be analogous to
the K3T1 behaviour found in the TNA–MeOϪ and related
systems. The regioselectivity could be designated K5T7, as
In view of the present low temperature NMR studies, the
inability to detect the O-bonded C-7 DNBFؒOPhϪ adduct, 3a,
in DMSO can be rationalized by considering the high reactivity
of DNBF towards carbon nucleophiles in DMSO-rich
media.11–14,16 For example, neutral electron-rich benzenes like
1,3,5-trimethoxybenzene or 3,5-dimethoxyaniline readily form
C-adducts with DNBF,13,14 and kinetic studies on C-attachment
of 3,4-diaminothiophene towards 125 have established a high
value for the forward rate constant, k1, of the order of ca.
9 × 105 dm3 molϪ1 sϪ1 in 50% H2O–50% DMSO. This value
compares very favourably with the forward rate coefficient for
O-attack on DNBF by hydroxide ion,19,20 which under similar
conditions has been determined to be k1 = 3.3 × 105 dm3
previously suggested 17 and
a qualitative energy–reaction
coordinate profile (based on the quantitative energy profiles
for the TNA–MeOϪ system)2b,30 proposed (Fig. 3). In agree-
ment with the designation, the forward rate constant for
attack at C-5 would be greater than that for attack at C-7, i.e.
k5 > k7. The greater thermodynamic stability attributed to the
C-7 adduct would require that the equilibrium constant for
formation of the C-7 adduct be larger than that for form-
ation of its C-5 counterpart; K7 > K5. Based on the TNA–
MeOϪ model system, it is also likely that the reverse rate
constants for decomposition of the adducts, kϪ5 and kϪ7
respectively, would bear the following relationship to one
another: kϪ5 > kϪ7
,
molϪ1
s
Ϫ1. Thus, the intrinsically low stability of phenoxide
.
oxygen-centred adducts6,17,22 that has been partly linked to the
superior leaving group ability of aryloxides,26 coupled with
the decreased activation barrier to C-adduct formation by
unhindered C-nucleophiles with 1, would account for our
inability to detect the O-adduct, 3a, in DMSO, even though its
formation is kinetically favoured (based on our low temperature
NMR study).
In the DNBF–ArOϪ systems, whether the investigation was
undertaken at room temp. in [2H6]DMSO as in the DNBF–3,5-
DTBPhOϪ case, or at low temperature in MeCN–DME as in
the DNBF–PhOϪ reaction, no C-5 adducts were observed.
Under the most sensitive experimental regime, only C-7
O-adducts were detected as the products of kinetic control.
Recall that with NBF, alkoxides do form C-5 adducts as kinetic
products that give way to their more stable C-7 analogues.1,23
Conversely, neither C-5 aryloxide O- nor C-adducts of NBF
were observed even at low temperature and it has been argued
that the corresponding C-7 O-adducts are the products of both
kinetic and thermodynamic control.9 Further, the literature
contains only one tentative report of formation of a transient
C-5 adduct formed by 1 and azide ion that gave way over
time in favour of a C-7 adduct; conversion of 1 to 2,4,6-
trinitroaniline was apparently also a competitive process in
this system.18 Apparently observation of a C-5 adduct in the
DNBF systems is rare, and when coupled with our inability
to observe C-5 → C-7 isomerization in the NBF–ArOϪ
systems, argues in favour of designating the regioselectivity
Classification of regioselectivity of phenoxide O-attack with 1
A full range of regioselectivity is exhibited in the reaction of
nucleophiles with 2,4,6-trinitroanisole and related picryl ethers:
K3T1 (TNA–MeOϪ), K1T1 (TNA–PhOϪ, where PhOϪ acts as
an O-nucleophile), K3T3 (TNA–PhOϪ, where PhOϪ acts as a
C-nucleophile) and K1T3 (TNAؒOMesϪ).22,26–29 Here K1T3
represents kinetically favoured formation of a C-1 adduct but
thermodynamic preference for C-3 adduct formation; the other
designations are derived similarly. In fact, these patterns of
regioselectivity have been found in a large number of systems
and have been reviewed recently.17
The classification of regioselectivity may be profitably
J. Chem. Soc., Perkin Trans. 2, 1997
1023

View Article Online
found in the current systems for both O- and C-aryloxide
nucleophiles as K7T7. Nonetheless, very rapid attack at C-5
and rapid rearrangement could mimic K7T7 behaviour and
make a K5T7 system appear to fit the K7T7 description
(dotted line in K5T7 profile, Fig. 3). Consequently, although we
strongly favour the classification of the regioselectivity in this
system as K7T7 we cannot rule out the possibility that it is
merely a ‘pseudo-K7T7’ system.
One factor that could account for both the kinetic and
thermodynamic preference apparently shown is the efficacy of
through-conjugation to the 4-nitro group (cf. structures 3–5).
Previously, this stabilization by through-conjugation was
advanced as a rationale for the K7T7 regioselectivity displayed
by aryloxide nucleophiles with NBF.9 Regardless of the struc-
ture of the electrophile, if a relatively later transition state (TS)
obtains for aryloxides26 as compared to alkoxides31 (based on
60% C᎐O bond formation between PhOϪ and TNA at C-126
relative to 45% C᎐O bond formation between MeOϪ and TNA
at C-131) then the TS for C-7 attack on DNBF should also
partake of the stabilization afforded by through-conjugation
to the 4-NO2 group. Thus, the TS for C-7 attack is stabilized
relative to the TS for C-5 attack and, at the same time, the C-7
adduct is stabilized relative to its C-5 counterpart. K7T7 regio-
selectivity results.
C-7 ortho-C and C-7 para-C adducts were found to form in a
statistical 2:1 ratio.7a
Although reaction via the ortho positions of phenoxide ion
as compared to para C-attack is favoured by statistics, it
should be re-emphasized that ortho and para sites are not
inherently equal in their potential nucleophilicity. In this con-
text, the 13C chemical shift of signal of the para carbon of
potassium phenoxide is 12.7 ppm upfield of that for the ortho
carbons in this nucleophile.10 This observation may be attrib-
uted to a greater shielding of the para carbon relative to the
ortho carbons as a result of the greater partial negative
charge density that resides at the para position 36 and which
could be expected to render the para site more nucleophilic
than a given ortho site. In this sense, a less than statistical
ratio for ortho/para attack could be expected in reactions of
phenoxide ion as a C-nucleophile.
Furthermore, steric factors play a role in determining
the feasibility of σ-adduct formation with ortho-substituted
benzenes37 and we have previously invoked increased steric
hindrance for phenoxide as an ortho nucleophile5,6 to account
partly for the observed preference for C-adduct formation via
the para carbon of PhOϪ. In the present case of the DNBF–
PhOϪ system, steric hindrance between the flanking NO2 group
and furoxan ring with phenoxide can either raise the rate
determining kinetic barrier to formation of the ortho quinoidal
intermediate, 4q (lower k1C, Scheme 2), or lead to a decrease in
the rate of tautomerization (lower k2C) that yields the ortho
Comparison of the current systems with the TNA–ArOϪ
systems previously examined bears consideration. In the TNA
systems it was deduced that the stability of C-1 adducts was
intimately tied to the degree of stereoelectronic stabilization
provided by the n–σ* interaction in these adducts, which are
geminally disubstituted with electronegative groups.6,22,27 On
the other hand, DNBF is unsubstituted at the C-7 position
and similar n–σ* stabilization of the resultant DNBFؒOPhϪ
adduct, 3a, is not possible. In fact, our studies of the NBF–
ArOϪ systems9 and the TNB–ArOϪ systems10a clearly show
that the preferred alignment of the attached aryloxy group at
C-7/C-2 in the aryloxide O-adduct is one which places the
p-type lone pair orbital of the oxygen coplanar with the aryl
ring of the attached aryloxyl moiety. This arrangement permits
conjugation (p–π) of the oxygen with the aryl ring and, unless
precluded by highly unfavourable steric congestion, dictates the
conformational preferences of the adduct.9,10a It is reasonable
to suggest that the preferred structures of the DNBF aryloxide
adducts reported herein also arise from this stabilizing p–π
stereoelectronic effect.§
C
C-adduct, 4. If the decrease in k2 is great enough, then the
process that results in decomposition of 4q back to 1 and PhOϪ
(kϪ1C) may even become competitive with the forward rate
process that yields adduct 4. By whatever mechanism steric
hindrance influences the rate of reaction for ortho C-attack as
compared to para C-reaction, the result is the same: nucleo-
philic attack via the para carbon becomes the favoured process
in the DNBF–PhOϪ system.
In conclusion, the current study of the regioselectivity of
aryloxide ambident O- and C-nucleophilic attack on DNBF
extends our understanding to super-electrophilic heterocycles
and shows the generality of the classification scheme (Fig. 3),
and is in accord with our analysis of stereoelectronic effects in
these systems.17
Experimental
Materials and methods
Ortho versus para C-attachment by phenoxide ion
4,6-Dinitrobenzofuroxan (DNBF, 1) was prepared by nitra-
tion 38 of benzofuroxan (Aldrich) and recrystallized from ethyl
acetate, mp 172 ЊC. [2H3]MeCN, [2H6]DMSO and [2H10]DME
(Merck or CDN) were dried by treatment with 4 or 3 Å molec-
ular sieves prior to use, as advocated by Burfield.39 1,4-
Dibromobenzene (DBB integral standard, Eastman) was
recrystallized from ethanol, mp 89 ЊC. Potassium ethoxide
(KOEt) solutions were prepared from freshly cut potassium
metal and dry EtOH (distilled from Mg turnings) under N2 and
standardized against potassium hydrogen phthalate (phenol-
phthalein indicator). Phenol (BDH) was distilled under vacuum
and stored and handled in an Ar-filled glovebox. 3,5-Di-tert-
butylphenol (Aldrich) was purified by recrystallization from
light petroleum. Melting points were measured on a Thomas-
Hoover capillary apparatus and were not corrected. Potassium
phenoxide (PhOK) and potassium 3,5-di-tert-butylphenoxide
(3,5-DTBPhOK) were prepared from the purified phenol and
standard KOEt and EtOH in a dry box; excess EtOH was
evaporated under a stream of N2 and the solid dried under
vacuum.
Formation of the para C-bonded C-7 DNBFؒPhOHϪ adduct, 5,
has been found to be favoured over formation of the ortho C-7
DNBFؒPhOHϪ adduct, 4, by a ratio of ca. 6:1 in the current
study. This observation is consistent with the findings of studies
of phenoxide C-attachment to TNB,5,10 where, again, formation
of the para-isomeric C-adduct is preferred. Interestingly, in our
previous investigations of phenoxide reactivity towards 2-(2Ј,4Ј-
dinitrophenyl)- and 2-(4Ј-nitrophenyl)-4,6-dinitrobenzotriazole
1-oxides (DNP–DNBT, 6, and NP–DNBT, 7, respectively) the
O–
Y
O2N
N
+
N
NO2
N
Z
NO2
Y = H, Z = NO2
Y = H, Z = H
6
7
NMR experiments
The NMR experiments were carried out on a Bruker AM-400
spectrometer (operating at 400.1 MHz) in [2H3]MeCN :
[2H10]DME (1:1 v/v) and in [2H6]DMSO. In the mixed solvent
§ We note the contribution by Professor N. S. Zefirov to the under-
standing of stereochemistry of stereoelectronic effects on the con-
formations of ring systems.33–35
1024
J. Chem. Soc., Perkin Trans. 2, 1997

View Article Online
1968, 22, 123; (h) E. Buncel, J. M. Dust and R. A. Manderville,
J. Am. Chem. Soc., 1996, 118, 2072.
system, residual CD2HCN served as chemical shift standard
(1H: δ 1.93 ppm) and lock signal, while spectra recorded in
[2H6]DMSO were referenced to the CD2HSOCD3 peak (δ 2.50
ppm). Chemical shifts are given in ppm; coupling constants are
reported in Hz. Wilmad PP-507 NMR tubes (5 mm) were used
in all experiments. Stock solutions and NMR tubes were
capped with rubber septa and swept out with N2 prior to injec-
tion of the reactants by means of a syringe.
3 (a) Ghosh and M. W. Whitehouse, J. Med. Chem, 1968, 11, 305; (b)
M. J. Strauss, A. DeFusco and F. Terrier, Tetrahedron Lett., 1981,
22, 1945.
4 (a) F. Sanger, Biochem. J., 1945, 39, 507; (b) J. F. Bunnett and
D. H. Hermann, Biochemistry, 1970, 9, 816; (c) M. Barra and
R. H. de Rossi, J. Org. Chem., 1989, 54, 5020; (d) J. M. Dust
and J. M. Harris, J. Polym. Sci., Part A.: Polym. Chem., 1990, 28,
1875.
5 (a) E. Buncel and W. Eggimann, J. Am. Chem. Soc., 1977, 99, 5958;
(b) E. Buncel, J. G. K. Webb and J. F. Wiltshire, J. Am. Chem. Soc.,
1977, 99, 4429; (c) E. Buncel and W. Eggimann, Can. J. Chem., 1976,
54, 2436; (d) E. Buncel, A. Jonczyk and J. G. K. Webb, Can. J.
Chem., 1975, 53, 3761; (e) E. Buncel and J. G. K. Webb, J. Am.
Chem. Soc., 1973, 95, 8470.
Representative room temperature experiment in DMSO.
Transfer of 100 µl of a [2H6]DMSO stock solution of DNBF
(1, 0.5 ) into an NMR tube that contained solvent (295 µl)
and DBB (5 µl from a 1  stock solution) afforded the initial
sample. DBB functioned as the internal integral standard and
was present in any experiment where the singlet for the aromatic
ring did not overlap with signals for adducts. Injection of 1
equiv. of 3,4-DTBPhOK (100 µl of a 0.5  stock solution) initi-
6 E. Buncel, J. M. Dust, A. Jonczyk, R. A. Manderville and I. Onyido,
J. Am. Chem. Soc., 1992, 114, 5610.
7 (a) E. Buncel and J. M. Dust, Can. J. Chem., 1988, 66, 1712;
(b) E. Buncel, R. A. Renfrow and M. J. Strauss, J. Org. Chem., 1987,
52, 488; (c) E. Buncel, J. M. Dust, K. T. Park, R. A. Renfrow and
M. J. Strauss, in Nucleophilicity, eds. J. M. Harris and S. P. McManus,
Adv. Chem. Ser. 215, ACS, Washington, D.C., 1987, pp. 369–378;
(d) R. A. Renfrow, M. J. Strauss, S. Cohen and E. Buncel, Aust. J.
Chem., 1983, 36, 1843; (e) E. Buncel and K. T. Park, in Physical
Organic Chemistry, 1986, ed. M. Kobayashi, Elsevier, Amsterdam,
1987, pp. 247–256.
8 (a) S. M. Shein and O. G. ByvalЈkevich, Zh. Org. Khim., 1972, 8, 328;
(b) S. M. Shein, O. G. ByvalЈkevich and A. D. Khemlinskaya, Zh.
Org. Khim., 1976, 12, 134; (c) V. Machacek, V. Sterba, A. Lycka and
D. Snobl, J. Chem. Soc., Perkin Trans. 2., 1982, 355.
9 R. A. Manderville and E. Buncel, J. Chem. Soc., Perkin Trans. 2.,
1993, 1887.
10 (a) R. A. Manderville and E. Buncel, J. Phys. Org. Chem., 1993, 6,
71; (b) F. Terrier, J. C. Halle, P. MacCormack and M. J. Pouet, Can.
J. Chem., 1989, 67, 503; (c) F. Terrier, R. Goumont, M. J. Pouet and
J. C. Halle, J. Chem. Soc., Perkin Trans. 2, 1995, 1629.
11 F. Terrier, E. Kizillian, J. C. Halle and E. Buncel, J. Am. Chem. Soc.,
1992, 114, 1740.
12 M. J. Strauss, R. A. Renfrow and E. Buncel, J. Am. Chem. Soc.,
1983, 105, 2473.
13 F. Terrier, J. C. Halle, M. J. Pouet and M. P. Simonnin, J. Org.
Chem., 1986, 51, 409.
1
ated the reaction. H NMR spectra were recorded at various
intervals but generally as rapidly as possible (i.e. within 3 min)
at the start of the reaction and then at progressively longer
intervals as the reaction proceeded. The system was typically
monitored until no further change could be detected in the
recorded spectrum.
In a separate experiment, acidification of the reaction mix-
ture (5 µl TFA) was performed after acquisition of an initial
spectrum. Spectra of the acidified solution were recorded
immediately following the addition of trifluoroacetic acid (with-
in 3 min).
Typical low temperature NMR experiment in MeCN–DME
(1:1 v/v). The PhOK stock solution (in [2H3]MeCN :[2H10]-
DME 1:1 v/v) was injected into an NMR tube and the solu-
tion then frozen by immersion in liquid N2. To this frozen
solution was injected 1 equiv. of DNBF solution (prepared in
the same solvent). The resultant frozen mixture (final volume:
500 µl; 0.1  in both components) was placed in a dry ice–
acetone bath, which had been maintained at Ϫ50 ЊC. Once
the contents of the tube had thawed at Ϫ50 ЊC, the com-
ponents were mixed by rapid inversion of the tube. The con-
tents were then frozen again by immersion of the tube in
liquid N2. The sample was transferred to the spectrometer
probe (Ϫ40 ЊC) and spectra were recorded at various inter-
vals. A standard sequence was: 3, 5, 7 and 9 min and then as
warranted by the observed changes in the acquired spectrum.
At the same time the temperature of the probe was gradually
raised to ambient.
14 J. C. Halle, M. P. Simonnin, M. J. Pouet and F. Terrier, Tetrahedron
Lett., 1985, 26, 1307.
15 R. Taylor, Electrophilic Aromatic Substitutions, Wiley, New York,
1990.
16 F. Terrier, J. C. Halle, M. P. Simonnin and M. J. Pouet, J. Org.
Chem., 1984, 49, 4363.
17 E. Buncel, J. M. Dust and F. Terrier, Chem. Rev., 1995, 95,
2261.
18 W. P. Norris, R. J. Spear and W. P. Read, Aust. J. Chem., 1983, 36,
297.
19 F. Terrier, F. Millot and W. P. Norris, J. Am Chem. Soc., 1976, 98,
5883.
Acknowledgements
20 F. Terrier, H. A. Sorkhabi, F. Millot, J. C. Halle and R. Schaal, Can.
J. Chem., 1980, 58, 1155.
21 E. Buncel, J. M. Dust, R. A. Mandervill, K. T. Park and I. Onyido,
Bull. Soc. Chim. Fr., 1988, 361.
22 R. A. Manderville and E. Buncel, J. Am. Chem. Soc., 1993, 115,
8985.
Financial support of this research from the NSERC (Canada)
(E. B.), the American Cancer Society (R. A. M.) and Sir Wilfred
Grenfell College (Principal’s Research Fund, J. M. D.) is grate-
fully acknowledged.
23 (a) E. Buncel, N. Chuaqui-Offermanns, B. K. Hunter and
A. R. Norris, Can. J. Chem., 1977, 55, 2852; (b) F. Terrier,
A. P. Charousse and F. Millot, J. Org. Chem., 1980, 45, 2666.
24 J. M. Dust and E. Buncel, Can. J. Chem., 1994, 72, 218.
25 F. Terrier, M. J. Pouet, E. Kizillian, J. C. Halle, F. Oururquin and
C. Paulmier, J. Org. Chem., 1993, 58, 4696.
26 C. F. Bernasconi and M. C. Muller, J. Am. Chem. Soc., 1978, 100,
5530.
27 E. Buncel, R. M. Tarkka and J. M. Dust, Can. J. Chem., 1994, 72,
1709.
References
1 (a) I. M. Sosonkin, A. Ya. Kaminsky, S. S. Gitis, V. A. Subbotin and
T. K. Polynnikova, Zh. Org. Khim., 1973, 9, 1470; (b) J. F. Bunnett,
Acc. Chem. Res., 1978, 11, 413; (c) R. Bacaloglu, C. A. Bunton and
G. Cerichelli, J. Am. Chem. Soc., 1987, 109, 621; (d) R. Bacaloglu,
A. Blasko, C. A. Bunton, F. Ortega and C. Zucco, J. Am. Chem.
Soc., 1992, 114, 7708.
2 (a) V. N. Drozd and V. N. Knyazev, Izv. Sib. Otd. Akad. Nauk
SSSR, Ser. Khim. Nauk., 1987, 52; (b) E. Buncel, M. R. Crampton,
M. J. Strauss and F. Terrier, Electron-deficient Aromatic- and
Heteroaromatic-base Interactions. The Chemistry of Sigma Anionic
Complexes, Elsevier, Amsterdam, 1984; (c) E. Buncel, The
Chemistry of Functional Groups. Supplement F. The Chemistry of
Amino, Nitro and Nitroso Compounds, ed. S. Patai, Wiley, London,
1982; (d) F. Terrier, Nucleophilic Aromatic Displacement. The
Influence of the Nitro Group, ed. H. Feuer, VCH, New York, 1991;
(e) F. Terrier, Chem. Rev., 1982, 82, 77; (f) G. A. Artamkina,
M. P. Egorov and I. P. Beletskaya, Chem. Rev., 1982, 82, 427;
(g) E. Buncel, A. R. Norris and K. E. Russell, Quart. Rev. London,
28 (a) K. L. Servis, J. Am. Chem. Soc., 1967, 89, 1508; (b) K. L. Servis,
J. Am. Chem. Soc., 1965, 87, 5495.
29 (a) R. A. Chamberlin and M. R. Crampton, J. Chem. Soc., Perkin
Trans. 2., 1995, 1831; (b) R. A. Chamberlin and M. R. Crampton,
J. Chem. Soc., Perkin Trans. 2., 1994, 425; (c) A. R. Norris,
Can. J. Chem., 1969, 47, 2895; (d) E. Buncel, A. R. Norris and
W. Proudlock, Can. J. Chem., 1968, 46, 2759.
30 C. F. Bernasconi, J. Am. Chem. Soc., 1970, 92, 4685.
31 C. F. Bernasconi and K. Howard, J. Am. Chem. Soc., 1982, 104,
7748.
32 R. A. Manderville, J. M. Dust and E. Buncel, unpublished work.
J. Chem. Soc., Perkin Trans. 2, 1997
1025

View Article Online
33 N. S. Zefirov, V. A. Samoshin, O. A. Subbotin, T. G. Zemlyanova
and G. M. Akhematova, Zh. Org. Khim., 1987, 23, 1320.
38 P. Drost, Justus Liebigs Ann. Chem., 1899, 307, 49.
39 D. R. Burfield and R. H. Smithers, J. Org. Chem., 1978, 43, 3966.
34 N. S. Zefirov, V. A. Palyulin and E. E. Dashevskaya, J. Phys. Org.
Chem., 1990, 3, 147.
35 N. S. Zefirov and V. A. Palyulin, Top. Stereochem., 1991, 20, 171.
36 (a) A. Julg and M. Bonnet, J. Chim. Phys., 1962, 59, 194;
(b) A. Pross, L. Radom and R. W. Taft, J. Org. Chem., 1980, 45, 818.
37 G. Doddi, G. Illuminati and F. Stegel, J. Org. Chem., 1971, 36, 1918.
Paper 6/05907D
Received 27th August 1996
Accepted 13th December 1996
1026
J. Chem. Soc., Perkin Trans. 2, 1997