Thermodynamic study of σH complexes in nucleophilic aromatic substitution reactions: Relative stabilities of electrochemically generated radicals

Article abstract of DOI:10.1002/ejoc.200701098

The mechanism for the electrochemical oxidation of σ^H complexes, such as 1-hydro-1-alkoxy/sulfoxy or -fluoro-2,4-dinitro/2,4,6- trinitrocyclohexadienyl anions, has been widely studied by means of cyclic voltammetry and controlled-potential electrolysis. Previous studies have shown that the electrochemical oxidation of σ^H complexes, formed by the addition of carbon or nitrogen nucleophiles followed by a two electron mechanism, corresponding to the formal elimination of the hydride anion (nucleophilic aromatic substitution of hydrogen mechanism, the NASH mechanism). For these σ^H complexes (Nu^- = OH^-, ^-OR, ^-SR, ^-F), the electrochemical reaction takes place by a one-electron mechanism and is followed by the radical elimination of the leaving group with the consequent recovery of the starting material. This mechanism is similar to that proposed for the electrochemical oxidation of σ^X complexes (nucleophilic aromatic substitution of a heteroatom, the NASX mechanism). The operating mechanism in each case, the NASH or NASX, can be rationalized in terms of thermodynamics. The standard potentials of the σ complex and/or the leaving group as well as the bond dissociation energies (BDEs) are determinant factors. This study has not led to a significant improvement in the electrochemical preparation of aromatic-substituted compounds, but does help to understand and predict the usefulness or uselessness of using the nucleophilic aromatic substitution route to obtain a desired product. Finally, the current approach extends the electrochemical methodology to different chemical fields, for example, to general nondestructive methods for the detection, identification, and quantification of either organic pollutants or explosives in different solvents. Wiley-VCH Verlag GmbH & Co. KGaA, 2008.

Full text of DOI:10.1002/ejoc.200701098

FULL PAPER
DOI: 10.1002/ejoc.200701098
Thermodynamic Study of σH Complexes in Nucleophilic Aromatic Substitution
Reactions: Relative Stabilities of Electrochemically Generated Radicals
Iluminada Gallardo*[a] and Gonzalo Guirado*[a]
Keywords: σH Complexes / Nucleophilic substitution / Aromatic substitution / Electrochemistry / Oxidation / Reaction
mechanisms
The mechanism for the electrochemical oxidation of σH com- a heteroatom, the NASX mechanism). The operating mecha-
plexes, such as 1-hydro-1-alkoxy/sulfoxy or -fluoro-2,4-dini-
tro/2,4,6-trinitrocyclohexadienyl anions, has been widely
studied by means of cyclic voltammetry and controlled-po-
tential electrolysis. Previous studies have shown that the
electrochemical oxidation of σH complexes, formed by the
addition of carbon or nitrogen nucleophiles followed by a two
electron mechanism, corresponding to the formal elimination
of the hydride anion (nucleophilic aromatic substitution of
hydrogen mechanism, the NASH mechanism). For these σH
complexes (Nu– = OH–, –OR, –SR, –F), the electrochemical re-
action takes place by a one-electron mechanism and is fol-
lowed by the radical elimination of the leaving group with
the consequent recovery of the starting material. This mecha-
nism in each case, the NASH or NASX, can be rationalized
in terms of thermodynamics. The standard potentials of the
σ complex and/or the leaving group as well as the bond
dissociation energies (BDEs) are determinant factors. This
study has not led to a significant improvement in the electro-
chemical preparation of aromatic-substituted compounds,
but does help to understand and predict the usefulness or
uselessness of using the nucleophilic aromatic substitution
route to obtain a desired product. Finally, the current ap-
proach extends the electrochemical methodology to different
chemical fields, for example, to general nondestructive meth-
ods for the detection, identification, and quantification of
either organic pollutants or explosives in different solvents.
nism is similar to that proposed for the electrochemical oxi- (© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
dation of σX complexes (nucleophilic aromatic substitution of
Germany, 2008)
complex (1) involves the loss of two electrons and a proton,
which corresponds to the formal elimination of a hydride
anion (Scheme 2, A).[6] Moreover, from an electrochemical
Introduction
The nucleophilic aromatic substitution (SNAr) of hydro-
gen, NASH, or of a heteroatom, NASX, constitutes an im-
portant step in numerous and important synthetic pro-
cedures, including the synthesis of herbicides, fungicides,
pesticides, pharmaceutical products, medicines, and numer-
ous antidepressants and antitumoral reagents,[1–5] so many
mechanistic studies have been performed.[1] The most ac-
cepted mechanism for SNAr is the addition/elimination
mechanism, a polar mechanism in which the first step is the
addition of an electron-rich compound (nucleophile) to an
electron-deficient aromatic one, leading to an anionic inter-
mediate commonly named a σ or Meisenheimer complex.
The reaction is completed in a second step, the departure
of a leaving group (Scheme 1).[1–4] The efficiency of the re-
action is strongly related to the nature of the leaving group,
so in the case of NASH processes, in which the hydride acts
as the leaving group, it will rarely be effective. However, it
is possible to electrochemically activate the NASH reaction.
The mechanism for the electrochemical oxidation of the σH
[a] Departament de Química, Univeristat Autònoma de Barcelona,
08193 Bellaterra, Barcelona, Spain
Fax: +00-34-93-581-29-20
Scheme 1. Nucleophilic aromatic substitution (SNAr) of hydrogen
(NASH) or of a heteroatom (NASX) following the addition/elimi-
nation mechanism.
E-mail: Iluminada.Gallardo@uab.es
Gonzalo.Guirado@uab.es
Eur. J. Org. Chem. 2008, 2463–2472
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2463

I. Gallardo, G. Guirado
FULL PAPER
Scheme 2. Mechanism for the electrochemical oxidation of A) 1,3-dinitrobenzene + (CH3)4NBH4 (σH complexes, the NASH process) and
B) 1-chloro-2,4-dinitrobenzene + (CH3)4NF (σX complexes, the NASX process).
point of view, the nucleophilic aromatic substitution of a have used carbon-centered nucleophiles (such as CN–, R–,
–
heteroatom (the NASX process) can also be electrochemi- and RCORCH2 , where R is mainly an alkyl group) as well
cally promoted.[7] The NASX electrochemical mechanism as nitrogenous ones, allowing a general method to be con-
also involves the elimination of a leaving group, although sidered for an electrochemically promoted NASH process.[6]
its departure is as a radical rather than as an anion This methodology, in principle, would be equally valid
(Scheme 2, B) as the rearomatization of the σX or ipso com- when oxygen, sulfur, or fluoride anions are used as nucleo-
plex (2) involves the loss of an electron. This information is philes as these are also capable of reacting with electron-
analogous to that obtained by combining other more tradi- deficient compounds, such as nitroaromatic ones, and,
tional techniques, such as UV/Vis and NMR spec- therefore, are able to form the corresponding anionic inter-
troscopy.[1-4, 8,9] Interestingly, the electrochemical oxidation mediates (Scheme 3). Some of these σH complexes have
1
of these σX complexes makes it possible to “jump” from been previously identified by UV/Vis (3–8)[11] or H NMR
the “polar” SNAr addition/elimination mechanism spectroscopy (5, 7, and 8)[12] in the course of kinetic studies,
(Scheme 1, σX complexes) to the “radical” heteroatom sub- although their chemical evolution after the chemical oxi-
stitution in aromatic compounds (Scheme 2, B).
dation process has never been studied. Hence, the aim of
The methodology employed satisfies the most important this work was not only to analyze the scope of electrochemi-
requirements of “Green Chemistry” processes, providing a cally promoted nucleophilic aromatic substitution, but also
new general environmentally friendly synthetic pro- to disclose the mechanism for the electrochemical oxidation
cedure.[10] The promising results that have been published of the intermediates. Note that in the case of these σH com-
Scheme 3. σH Complex structures.
2464
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 2463–2472

σH Complexes in Nucleophilic Aromatic Substitution Reactions
plexes derived from the attack of oxygen, sulfur, or fluorine σ complex is 1,3-DNB, which indicates that the σ complex
anions, two mechanisms can potentially take place. The present in the mixture is the σH complex 3. In this case the
first, the NASH process, would involve the breaking of the CH3O group is lost rather than the hydrogen, which would
C–H bond in what would be an attractive process from a have been lost under the NASH mechanism (Scheme 2).[13]
synthetic point of view, whereas the second would involve These results were confirmed by oxidative controlled-poten-
the rupture of the C–X bond, similar to the NASX process, tial electrolysis of the reaction mixture when 100% σ com-
that is, a radical mechanism similar to that described above plexes were produced. The initial reagent was mainly reco-
for σX complexes. Thus, the initial reactant would be reco- vered, 1,3-DNB (95%), and only 5% of the substituted
vered following an interesting mechanistic process.
product, 3-nitroanisole, was obtained after the passage of
1 F at 1.00 V. In none of the analyses of the oxidized mix-
tures was 2,4-dinitroanisole, which is the expected product
for the NASH mechanism, even detected (Table 1).
Results and Discussion
Cyclic voltammetry experiments show that after each ad-
dition of a hydroxide nucleophile solution it is possible to
observe in an anodic scan from 0.00 to 1.50 V the presence
of two new peaks at 0.59 and 1.21 V (vs. SCE), respectively
(see the Exp. Sect.). The current anodic peak values increase
with the concentration of the added nucleophile. When a
pure solution of TBAOH (tetrabutylammonium hydroxide)
was analyzed by cyclic voltammetry, the presence of an oxi-
dation peak was revealed at 1.21 V (vs. SCE). Thus, analo-
gously to the experiments described above, it is possible to
assign the anodic peak at 0.59 V to the oxidation of the
formed σH complex (1-hydroxy-2,4-dinitrocyclohexadienyl
anion, 4). After the addition of 22.5 µL of the TBAOH
solution (1.4 equiv.), it was possible to observe: 1) The com-
plete disappearance, in a cathodic scan, of the first re-
duction peak of the initial nitroaromatic compound, 1,3-
DNB, 2) that the oxidation wave assigned to the σH com-
plex 4 at 0.59 V is a one-electron irreversible wave, and 3)
that if a second cathodic scan was recorded after the oxi-
dation counter part of the first scan, from 0.00/–1.00/1.50
to –1.00 V, the appearance of a reduction peak was ob-
served, which corresponds to the product formed after the
oxidation of the σH complex. In this case, the oxidation
peak value is exactly the same as that previously determined
for 1,3-DNB (starting material), as shown in Figure 1 for
the methoxy adduct. Thus, for oxygenated adducts, it seems
that the “oxidation” product obtained is the initial reagent
rather than the substitution product (2,4-dinitrophenol), in
contrast to the previously described examples of oxidation
of σH complexes in which a NASH process takes place.
These observations imply that the departure of the hydrox-
ide group, which had been previously introduced, is electro-
chemically promoted. In order to confirm these voltam-
metric results, electrolysis of the mixture was performed.
The electrochemical analysis performed by using cyclic
voltammetry of a DNB/MeOK mixture [1,3-dinitrobenzene
(DNB)/potassium methoxide (MeOK), see the Exp. Sect.]
showed the appearance of two new peaks at 0.72 and 1.33 V
(vs. SCE) in the oxidation scan after the addition of potas-
sium methoxide (Figure 1). A pure solution of potassium
methoxide was analyzed by cyclic voltammetry and showed
a one-electron oxidation peak at 1.33 V vs. SCE under the
same experimental conditions. Thus, these two oxidation
waves can be assigned to a σ complex and to the electro-
chemical response of the “free” methoxide anion present in
the mixture, respectively.
Figure 1. Cyclic voltammogram of 10 m 1,3-DNB in anhydrous
DMF under argon containing 0.1  TBABF4 as a supporting elec-
trolyte after the addition of 10 equiv. of potassium methoxide (T =
13 °C). The scan range: 0.00/–1.00/+1.50/–1.00/0.00 V (two cycles);
scan rate: 1 Vs–1; working electrode: glassy carbon disk with a dia-
meter of 0.5 mm.
Note that after the addition of 10 equiv. of methoxide After the passage of 1 F at 1.00 V, the analysis of the non-
the reduction peak corresponding to the starting material, colored solution by GC–MS confirmed the recovery of 1,3-
1,3-DNB, in a first cathodic scan totally disappeared (Fig- DNB in a quantitative yield.
ure 1). Thus, as no “free” reactant was detected, it is pos-
When an excess of sodium thiosulfate was added to the
sible to conclude that the nucleophilic attack corresponds 1,3-DNB solution (molar ratio 10:1), the presence a new
to 100%, so the σ complex is formed quantitatively. Under wide oxidation wave was observed at 0.47 V (vs. SCE) (see
these experimental conditions the electrochemical oxidation the Exp. Sect.). This should mainly be attributed to the
corresponds exclusively to the formation of the σ complex presence of the corresponding σ complex and to excess of
owing to the detection of a one-electron oxidation peak at the thiosulfate anion [Epa = 0.44 V (vs. SCE) at 1.0 Vs–1].
0.72 V in the anodic scan. As can be deduced from either After exhaustive controlled-potential electrolysis, the DNB
a cathodic counter or after two cycles, the electrochemical was not the only product recovered. The electrochemical
response of the product obtained from the oxidation of the oxidation also yielded two more products in moderate and
Eur. J. Org. Chem. 2008, 2463–2472
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2465

I. Gallardo, G. Guirado
FULL PAPER
Table 1. Oxidative electrolysis of σH complexes.
good yields: Ethyl 3-nitrophenyl thioether (14%) and recovery of the starting material indicates the σH complex
thiophenol (64%). These two new products were obtained 5, whereas the nitrophenylthio derivatives suggest the exis-
by NASX processes. Controlled-potential electrolysis and tence of a σX complex derived from the attack of the nu-
analysis of the product mixture allowed the type of σ com- cleophile on a substituted position leading to the substitu-
plexes and also their concentrations to be determined. The tion of the nitro group NO2. The fact that only one oxi-
2466
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 2463–2472

σH Complexes in Nucleophilic Aromatic Substitution Reactions
dation wave was obtained in the cyclic voltammogram can
be explained by the oxidation potential values of the σ com-
plexes present in the mixture. The anodic peak potential
value of the σH complex 5 is expected to be between 0.5–
0.7 as the molecule has two nitro groups. Similarly, the oxi-
dation peak value of the σX complex is expected to be of
the same potential range owing to the oxidation potential
–
of the leaving group (NO2 ). As a consequence, only one
oxidation peak was detected by cyclic voltammetry. The
higher yield of the 3-nitrothiophenol derivative can also be
justified by the experimental conditions applied. The first
product obtained during the electrolysis is the correspond-
ing ethyl thioether. However, as the electrolysis was exhaus-
tive this product can be oxidized leading to its radical cat-
ion. This radical cation is not a stable intermediate, and
rupture of the S–C bond, corresponding to the alkyl chain,
takes place. As a result, the final product is the correspond-
ing thiophenyl derivative (Table 1). Finally, note that no
products from aromatic hydrogen substitution were de-
tected.
Figure 2. Cyclic voltammetry of 10 m 1,3,5-trinitrobenzene
(TNB) in anhydrous DMF under argon containing 0.1  TBABF4
as a supporting electrolyte after the addition of 5 equiv. of anhy-
drous tetramethylammonium fluoride (T = 13 °C). The scan range:
0.00/–1.00/+1.50/–1.00/0.00 V; scan rate: 1 Vs–1; working electrode:
glassy carbon disk with a diameter of 0.5 mm.
The last type of nucleophile tested in these studies was stitution products were detected; in contrast, in our pre-
the fluoride anion. Fluoride solution was added until 1,3- viously published studies of σH complexes derived from car-
DNB was no longer detected by cyclic voltammetry. Thus, bon- or nitrogen-centered nucleophiles, a two-electron oxi-
all the nitroaromatics present in the mixture were derived dation process was observed (NASH mechanism). However,
from the σH complex. After an exhaustive oxidation of the in some cases the results shown in Table 1 reveal the syn-
reaction mixture (1 F), the only product obtained was the thetic utility of this reaction. An oxidative nitro group sub-
initial nitroaromatic reagent as a consequence of the oxi- stitution in 1,3-DNB and TNB by SEt– occurs with prepar-
dation of the σH complex 6. Thus, the only oxidation prod- ative yields that range from modest to good when a huge
uct detected was a consequence of the loss of fluorine rather excess of the nucleophile is used. The nucleophilic displace-
than hydrogen (Table 1).
ment of a nitro group activated by ortho or para functions
Figure 2 shows the cyclic voltammogram of 1,3,5-tri- other than nitro groups has been demonstrated in the chem-
nitrobenzene (TNB) in anhydrous DMF containing 0.1  ical literature,[15] but only a few synthetically useful SNAr
of TBABF4 after the addition of 1 equiv. of tetramethyl- reactions involving the displacement of the NO2 group in
ammonium fluoride. In the initial cathodic scan, it is pos- 1,3-DNB have been reported.[15]
sible to observe the remaining TNB that has not been at-
These results led us to ask why the same type of σH com-
tacked by the fluorine (Epa = –0.53 V vs. SCE),[14] so it was plexes yield different substitution products, and conse-
possible to calculate the percentage of the nucleophilic at- quently the different electrochemical oxidation pathways.
tack. In the corresponding anodic counter part two oxi- This should be strongly connected to the ability of the sub-
dation peaks at 0.26 and 1.09 V are observed. The first peak stituent as a leaving group and to the bond dissociation
is associated with the reactivity of the TNB anion radical, energies (BDEs) of the corresponding C–Nu or C–H bond.
as recently demonstrated, whereas the second one at 1.09 V The relationship between the thermodynamic properties of
corresponds to the oxidation of the σH complex 7 formed σH complexes and their radicals are conveniently schema-
by the attack of the fluoride nucleophile on TNB.[14] These tized in the thermochemical mnemonic shown in Scheme 4
results were easily confirmed by analysis of a reaction mix- for the NASX and NASH processes. As the key step in both
ture after controlled-potential electrolysis at 1.20 V (1 F) processes is the evolution of the σ complex radical (∆1G°),
when 100% of 7 was formed; the main product obtained comparison of the Gibbs energy values of this step would
was the starting nitroaromatic compound (Table 1).
indicate which process is operating in each case. It is re-
When either oxygen or sulfur nucleophiles were added to markable that, as expected, the value of ∆1G° for the NASX
a solution of TNB until the quantitative formation of their process is related to the standard potential of the leaving
corresponding σH complexes 8 and 9, respectively, a one- group and the corresponding σH complex formed. How-
electron wave was observed at around 1.00 V. The results ever, the ∆1G° value for the NASH process also involves the
obtained after controlled-potential electrolysis were very standard potential values of H·/H+ and the corresponding
similar to those described above for the DNB derivatives.
nitroaromatic compound. Moreover, the difference in BDEs
The results obtained in these studies with different types between a C–H and a C–Nu bond should also be taken
of nucleophiles and nitroaromatic compounds are summa- into account before analyzing the viability of the NASH
rized in Table 1. In the case of σH complexes derived from process. The overall results obtained for different σH com-
oxygen-, sulfur-, or fluorine-centered nucleophiles, a one- plexes and a σX complex are summarized in Table 2. For
electron oxidation wave was observed and no hydrogen sub- both processes, the values corresponding to the oxidation
Eur. J. Org. Chem. 2008, 2463–2472
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2467

I. Gallardo, G. Guirado
FULL PAPER
of the leaving group, E°Nu·/Nu–, have previously been re- the C–H and C–Nu bonds in relation to a benzene ring
ported.[16] However, no E° values have been measured for have also been published.[19–21] Note that the enthalpies
the σH complexes; oxidation peak potential values are used (BDEs) are considered rather than the standard free energy
as E° values. Fortunately, this will correspond to a system- of cleavage (∆G°, the driving force for bond cleavage) be-
atic and unaffected error as these values would not be cause the ∆S° value (the entropy for C–H or C–Nu cleav-
meaningfully different in these cases.[6,7,17] Moreover, for age) corresponding to the formation of two molecules from
the NASH process, the E°H+/H· and E°(nitroaromatic/nitro- one is small and does not vary significantly in the series
aromatic anion radical) values are, respectively, –1.53 and and may be approximately equal to 1 meVK–1
–0.83 V, which are the most accepted potentials for both (0.02 kcalmol–1).[22] Moreover, the entropic terms corre-
substances in organic nonprotic solvents, as has previously sponding to products and reactants can be simplified from
been published.[6a,12a,18] Finally, the BDE values used for the equation shown in Scheme 4 as the difference between
Scheme 4. Thermochemical mnemonic describing the NASX and NASH processes.
2468
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 2463–2472

σH Complexes in Nucleophilic Aromatic Substitution Reactions
two standard free energies can be expressed in enthalpy and
Note that the ∆1G° values shown in Table 2 do not in-
entropy terms, the entropic terms having the same values clude ∆3G°, which corresponds to ∆G° for the formation of
with the opposite sign. For reasons of comparison, the an- the σH complexes with the nucleophile. Most of these values
ionic intermediates 1, 9, and 10, which undergo NASH pro- are unknown with the exception of the thioethoxide, for
cesses,[6] are included in Table 2. For the same purpose, the which ∆3G° is –6.33 kcalmol–1.[23] However, as can be seen
σX complex 2 is also included as its electrochemical behav- from the NASX and NASH thermodynamic equations,
ior[7] is well known and it can be used as a reference for the ∆3G° is equally involved in both processes, so the calculated
NASX process.
values of ∆1G° will not be meaningfully different in terms
Table 2. Thermodynamic data.
[a] E [V] vs. SCE. [b] Epa [V] vs. SCE at 1 Vs–1. [c] In kcalmol–1. [d] Data from ref.[18]. [e] Data from ref.[6a]. [f] Data from ref.[19]. [g]
Energetically favored process is indicated in bold. [h] Data from ref.[17]. [i] Data from ref.[21]. [j] Data from ref.[20]. [k] ∆1G° is a an absolute
value as it includes the ∆3G° value from ref.[23]. [l] Data from ref.[25]. [m] Data from ref.[26]. [n] Data from ref.[24]
.
Eur. J. Org. Chem. 2008, 2463–2472
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2469

I. Gallardo, G. Guirado
FULL PAPER
Scheme 5. Mechanisms for the electrochemical oxidation of σH complexes.
of the relative energies of both processes. In all cases, ∆1G° electrochemical oxidation of σX complexes (the NASX
(NASX) is more favorable than ∆1G° (NASH). Although mechanism). The “NASX-type” mechanism for these σH
–
–
–
–
the average differences are around 25 kcalmol–1, they are complexes (Nu– = OH, OR, SR, F) contrasts with the
well outside the experimental errors of the standard poten- NASH mechanism, a two-electron mechanism that corre-
tial (10 mV, ca. 0.2 kcalmol–1) and the BDEs (2 kcalmol–1). sponds to the formal elimination of a hydride anion when
Note that when the SNAr reaction involves either nitro- H–, CN–, RCORЈ, and RNH2 are used as nucleophiles. The
gen- or carbon-centered nucleophiles, the experimentally two electrochemical oxidation pathways are compared in
observed mechanism corresponds to the NASH process. In Scheme 5.
this sense, the ∆1G° values for the NASX and NASH
Thus, the quantitative formation of σH complexes 3–9
mechanisms confirm that the second mechanism is the ther- does not yield the substitution product after an electro-
modynamically favored one (σH complexes 10 and 11, chemical oxidation process. The nature of the nucleophile
Table 2). Remarkably, in the case of σX complexes such as is a determinant factor. A thermodynamic study explains
2, we previously reported that the electrochemically pro- the operating mechanism (NASH or NASX) in terms of
moted SNAr reaction involves the departure of a chlorine standard potentials as well as BDE values by comparison
radical rather than a fluorine radical.[7a] This result can be of the relative stabilities of the σH complex radicals gener-
easily explained in terms of the standard potential of the ated by electrochemical oxidation. Although this study has
leaving group. As reported, the oxidation of F– is at least not led to any significant improvement in the electrochemi-
0.2 V more difficult than that of Cl–, E°F·/F–,DMF=2.59 and cal preparation of aromatic-substituted compounds, with
1.79 V at 298 K, respectively. Thus, ∆1G° in the NASX pro- the exception of the use of sulfur nucleophiles to displace a
cess is higher for the loss of fluorine, which effectively con- nitro group, it does enable a better understanding of the
firms that the departure of the leaving group from σX com- experimental results. Finally, the current approach extends
plexes is also thermodynamically controlled.
the electrochemical methodology to different chemical
Finally, fluorinated compounds have not been considered fields, for example, to general nondestructive methods for
because the literature BDEs for the aromatic fluorinated the detection, identification, and quantification of either or-
compounds vary from 87 to 127 kcalmol–1. The peculiar ganic pollutants[27] or explosives[28] in different solvents.
nature of F– led to larger discrepancies in the derived poten-
tial as no direct experimental electrochemical data is avail-
able. Hence, the standard potential for the oxidation of F–,
E°F·/F–,DMF at 298 K, is estimated to be between 2.59 and
Experimental Section
Chemicals: Anhydrous acetonitrile (ACN) and N,N-dimethylform-
2.62 V from the available values of the free energy of trans-
amide (DMF) stored in an inert atmosphere and molecular sieves
fer from water to DMF.[15,24] Clearly, this major uncertainty
were purchased from Across and SDS. Nitro aromatic compounds
and nucleophiles of the highest available purity were purchased.
affects the use of the aforementioned thermodynamic cycles
in estimating which process will operate in these cases.
Synthesis of σH Complexes with Oxygen-, Sulfur-, and Fluorine-Cen-
tered Nucleophiles
Conclusions
Potassium Methoxide (MeOK): Different controlled amounts of
potassium methoxide, which was used as a nucleophile, were added
to a DMF solution (5 mL) of 1,3-DNB (10 m) containing 0.1 
The electrochemical oxidation of σH complexes, such as
1-hydro-1-alkoxy/sulfoxy or fluoro-2,4-dinitro/2,4,6-tri-
TBABF4 (tetrabutylammonium tetrafluoroborate) as the support-
nitrocyclohexadienyl anions, occurs by a one-electron
ing electrolyte. The reaction mixtures were simultaneously analyzed
mechanism followed by the radical elimination of the leav-
ing group with the consequent recovery of the starting ma-
by UV/Vis spectrophotometry and cyclic voltammetry after each
addition. The solution immediately became colored (λmax
=
terial. This mechanism is similar to that proposed for the 578 nm), the blue color becoming darker as the concentration of
2470 Eur. J. Org. Chem. 2008, 2463–2472
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

σH Complexes in Nucleophilic Aromatic Substitution Reactions
the methoxide increased in the solution mixture. These general
spectroscopic features are characteristic of the formation of the σ
complex.[1] After the addition of 10 equiv. of potassium methoxide,
intermediate 3 was obtained quantitatively.
voltammograms were displayed on a Tektronix (2212) instru-
ment.[6]
Solutions were prepared using N,N-dimethylformamide (DMF) as
solvent and were purged with argon before the measurements. Ar-
gon was allowed to flush the solutions during the measurements.
The concentrations of the nitroaromatic compounds were around
10–3 ; 0.1  tetrabutylammonium tetrafluoroborate was used as
the supporting electrolyte.
Tetrabutylammonium Hydroxide (TBAOH): A solution of 5 m of
1,3-DNB in DMF (5 mL, 0.1  TBABF4) was prepared under ar-
gon. Later, controlled aliquots of a 1.5  TBAOH solution were
carefully added. The voltammetric responses were registered after
every addition of the nucleophile solution. Before the electrochemi-
cal measurements were performed, and after the addition of the
TBAOH aliquots, the solution turned red-colored, which is indica-
tive of the formation of σ complexes in solution. Thus, the presence
of these complexes was first confirmed by UV/Vis spectroscopy;
two characteristic absorption peaks for the σ complexes at 504 and
540 nm were clearly seen. After the addition of 1.4 equiv. of
TBAOH, intermediate 4 was obtained quantitatively.
Acknowledgments
We gratefully acknowledge the financial support of the Spanish
Ministerio de Educación
y Ciencia (MEC) through project
CTQ2006-01040.
[1] F. Terrier, Nucleophilic Aromatic Displacement (Ed.: H. Feuer),
VCH Publishers, New York, 1991, p. 257.
[2] J. March, Advanced Organic Chemistry, 4th ed., Wiley, New
York, 1992, p. 641.
Sodium Thiosulfate: A similar approach to those described above
were also applied to the sulfur nucleophile, sodium thiosulfate, and
the same general trends were observed. After the addition of
10 equiv. of sodium thiosulfate to a solution of 10 m 1,3-DNB in
DMF (5 mL), intermediate 5 was obtained quantitatively.
[3] E. Buncel, J. M. Dust, F. Terrier, Chem. Rev. 1995, 95, 2261.
[4] O. N. Chupakhin, V. N. Chupakhin, H. C. Van der Plas, Nucle-
ophilic Aromatic Substitution of Hydrogen, Academic Press,
New York, 1994.
[5] K. A. Parker, C. A. Coburn, J. Org. Chem. 1992, 57, 97.
[6] a) I. Gallardo, G. Guirado, J. Marquet, Chem. Eur. J. 2000, 7,
1759; b) I. Gallardo, G. Guirado, J. Marquet, patent pending
ES2000/489; c) I. Gallardo, G. Guirado, J. Marquet, Eur. J.
Org. Chem. 2002, 2, 251; d) I. Gallardo, G. Guirado, J. Mar-
quet, Eur. J. Org. Chem. 2002, 2, 261; e) I. Gallardo, G. Guir-
ado, J. Marquet, J. Org. Chem. 2003, 68, 7334.
Tetramethylammonium Fluoride: An anhydrous tetramethylammo-
nium fluoride solution in DMF under argon was added to 1,3-
DNB and 1,3,5-TNB solutions prepared under the same condi-
tions. New oxidation waves at 0.70 and 1.09 V, corresponding to
the formation of the σH complexes, were observed. The fluoride
solution was added until the 1,3-DNB was no longer detected by
cyclic voltammetry at which point all the nitroaromatic present in
the mixture had been converted to the σH complex.
[7] a) I. Gallardo, G. Guirado, J. Marquet, J. Org. Chem. 2002,
67, 2548; b) I. Gallardo, G. Guirado, J. Marquet, J. Elec-
troanal. Chem. 2000, 488, 64.
General Procedure for the NASX of σH Complexes in Nitroarenes:
A solution of nitroarene (10–20 m) in DMF (5 mL) containing
NBu4BF4 (0.1646 g, 0.1 ) as the supporting electrolyte was pre-
pared under nitrogen. The corresponding σH complex was prepared
by careful addition of the nucleophile to the solution of the nitroar-
ene under nitrogen. The oxidation peak potentials of the σH com-
plexes were measured by cyclic voltammetry. Then electrolysis was
carried out at values of potentials around 100 mV, which were more
positive than the values measured for each σH complex, using a
carbon graphite electrode as a working electrode. The electrolysis
was stopped when the starting material had totally reacted. Then
the mixture was extracted with water/toluene. The organic layer
was dried with Na2SO4 and evaporated to afford a residue that was
analyzed by gas chromatography. The analysis showed the presence
of nitro compounds. The final products were analyzed by GC, GC–
[8] a) M. R. Crampton, V. Gold, J. Chem. Soc. B 1967, 23; b)
M. R. Crampton, B. Gibson, J. Chem. Soc. Perkin Trans. 2
1981, 533; c) M. R. Crampton, R. Chamberlin, J. Chem. Re-
search. (S) 1967, 23; d) M. R. Crampton, S. D. Lord, J. Chem.
Soc. Perkin Trans. 2 1997, 369; e) M. R. Crampton, J. Delaney,
L. C. Rabbit, J. Chem. Soc. Perkin Trans. 2 1999, 2573.
[9] V. Machacek, V. Sterba, A. Lycka, D. Snobl, J. Chem. Soc.
Perkin Trans. 2 1982, 355.
[10] M. A. Matthews, Pure Appl. Chem. 2001, 73, 1305.
[11] a) For compound 3, see: F. Millot, F. Terrier, Bull. Soc. Chim.
Fr. 1971, 11, 3897; b) for compound 4, see: C. F. Bernasconi,
J. Am. Chem. Soc. 1970, 92, 4682; c) for compound 5, see:
M. R. Crampton, J. A. Stevens, J. Chem. Soc. Perkin Trans. 2
1991, 7, 925; d) for compound 6, see: J. H. Clark, M. S. Robert-
son, A. Cook, C. Streich, J. Flourine Chem. 1985, 28, 161; e)
for compound 7, see: M. P. Egorov, G. A. Artamkina, I. P. Be-
letskaya, O. A. Reutov, Izves. Akadem. Nauk SSSR, Ser. Chim.
1978, 10, 2431; f) for compound 8, see: E. Buncel, N. Chuaqui-
Offermanns, R. Y. Moir, R. A. Norris, Can. J. Chem. 1979, 57,
494.
[12] a) R. Chamberlin, M. R. Crampton, J. A. Stevens, J. Chem.
Soc. Perkin Trans. 2 1993, 1, 75; b) F. Terrier, G. Ah-Kow, M.
Pouet, M. P. Simmonnin, Tetrahedron Lett. 1976, 3, 227; c) V.
Machacek, A. Lycka, J. Kavalek, Magn. Reson. Chem. 2000,
38, 1001.
[13] a) C. P. Andrieux, J.-M. Savéant, Electrochemical Reactions in
Investigation of Rates and Mechanism of Reactions in Tech-
niques of Chemistry, vol. 6 (Ed.: C. F. Bernasconi), Wiley, New
York, 1986, chapter 2.1; b) C. P. Andrieux, Pure Appl. Chem.
1994, 66, 2445; c) simulations were performed by using DIGI-
SIM software, commercially available from BAS Corp.
[14] a) I. Gallardo, G. Guirado, J. Marquet, N. Vilà, Angew. Chem.
Int. Ed. 2007, 46, 1321; b) I. Gallardo, G. Guirado, J. Marquet,
Chem. Commun. 2002, 2638; c) N. A. Macías-Ruvalcaba, J. P.
Telo, D. H. Evans, J. Electroanal. Chem. 2007, 600, 294.
1
MS, H NMR, and cyclic voltammetry and identified by compar-
ing their spectroscopic behavior with either that reported in the
literature or with pure samples in each case. The product yields
were not optimized and were calculated by GC after verifying by
1H NMR of the crude and cyclic voltammetry that only the substi-
tution products and the starting material were present.
Cyclic Voltammetry: An electrochemical conical cell equipped with
a methanol jacket, which made it possible to fix the temperature at
13 °C by means of a thermostat, was used for the set-up of the
three-electrode system. For cyclic voltammetry experiments, the
working electrode was in all cases a glassy carbon disk with a dia-
meter of 0.5 mm. It was polished using a 1 µm diamond paste. The
counter electrode was a Pt disk with a diameter of 1 mm. All the
potentials were reported versus an aqueous saturated calomel elec-
trode (SCE) isolated from the working electrode compartment by
a salt bridge. The cyclic voltammetry apparatus was composed of
a home-made solid-state amplifier potentiostat with positive feed-
back iR drop compensation and a Tacussel GSTP 4 generator. The
Eur. J. Org. Chem. 2008, 2463–2472
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2471

I. Gallardo, G. Guirado
FULL PAPER
[15] a) N. Kornblum, L. Cheng, R. C. Kerber, M. M. Kestner, B. N.
Newton, H. W. Pinnick, R. G. Smith, P. A. Wade, J. Org. Chem.
1974, 41, 1560; b) J. R. Beck, Tetrahedron 1978, 34, 2057.
[16] C. P. Andrieux, J.-M. Savéant, A. Tallec, A. Tardivel, C. Tardy,
J. Am. Chem. Soc. 1997, 119, 2420.
[17] a) The differences between the peak potential and standard
oxidation potential[6,7] in the case of σH complexes is 0.03 V
(ca. 1 kcalmol–1), whereas in the case of σX complexes it is
0.08 V (ca. 2 kcalmol–1). Thus, these differences would not in-
troduce significant uncertainties into the determined values of
∆G°; b) W. Z. Liu, F. G. Bordwell, J. Org. Chem. 1996, 61,
4778.
[22] C. P. Andrieux, M. Farriol, I. Gallardo, J. Marquet, J. Chem.
Soc. Perkin Trans. 2 2002, 985.
[23] J. W. Larsen, K. Amin, S. Ewing, L. L. Magid, J. Org. Chem.
1972, 37, 3857.
[24] C. P. Andrieux, C. Combellas, F. Kanoufi, J.-M. Savéant, A.
Thirbault, J. Am. Chem. Soc. 1997, 119, 9527.
[25] J.-M. Saveant, J. Am. Chem. Soc. 1992, 114, 10595.
[26] J. Zhao, X. Cheng, X. Yang, J. Mol. Struct. (THEOCHEM)
2006, 766, 87.
[27] A. Akata, M. D. Gurol, Chem. Oxidation 1994, 2, 140.
[28] a) A. Hilmi, J. H. T. Luong, A.-L. Nguyen, Anal. Chem. 1999,
71, 873; b) K. Spiegel, T. Welsch, J. Fresenius, Anal. Chem.
1997, 357, 333.
[18] D. D. M. Wayner, V. D. Parker, Acc. Chem. Res. 1993, 26, 287.
[19] R. D. J. Froese, K. Morokura, J. Phys. Chem. A 1999, 103,
4580.
Received: November 20, 2007
Published Online: April 4, 2008
[20] Y. R. Luo, J. L. Homes, J. Phys. Chem. 1994, 98, 303.
[21] W. van Scheppingen, E. Dorrestijn, I. Arends, P. Mulder, J.
Phys. Chem. A 1997, 101, 5404.
2472
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 2463–2472