Electrochemical synthesis of organophosphorus compounds through nucleophilic aromatic substitution: Mechanistic investigations and synthetic scope

Article abstract of DOI:10.1002/ejoc.201101357

Advantages of the electrochemical approach in the nucleophilic aromatic substitution reaction, such as (a) low cost and ready availability of reagents, (b) atom economy, and (c) high yields (approaching 100 %), are applied to rationalize the (polar or radical) mechanism and to develop new greener synthetic routes for the synthesis of substituted nitroaromatic organophosphorus compounds. The nucleophiles used to study the feasibility and viability of the reaction are the classical tervalent phosphorus nucleophiles: trimethylphosphane, triethylphosphane, triphenylphosphane, diphenylphosphane, trimethyl phosphite, triethyl phosphite, dimethyl phosphonate, diethyl phosphonate, oxo(diphenyl)phosphorane, with two nitroaromatic compounds 1,3,5-trinitrobenzene and 1-chloro-2,4,6-trinitrobenzene in a DMF solution containing 0.1 M tetrabutylammonium tetrafluoroborate. In all cases, in order to establish the feasibility or benefits of the electrochemical approach relative to the chemical approach, blank reactions were also performed. Electrochemical oxidation of σ^H complexes and zwitterionic complexes obtained by nucleophilic attack of phosphorus nucleophiles was studied by means of cyclic voltammetry and controlled-potential electrolysis. The oxidation mechanism of those intermediates was disclosed, and the synthetic scope of the reaction was explored. Copyright

Full text of DOI:10.1002/ejoc.201101357

FULL PAPER
DOI: 10.1002/ejoc.201101357
Electrochemical Synthesis of Organophosphorus Compounds through
Nucleophilic Aromatic Substitution:
Mechanistic Investigations and Synthetic Scope
Hugo Cruz,[a] Iluminada Gallardo,*[a] and Gonzalo Guirado*[a]
Keywords: Nucleophilic substitution / Phosphorus nucleophiles / Electrochemistry / Nitroaromatic compounds /
Zwitterions
Advantages of the electrochemical approach in the nucleo-
philic aromatic substitution reaction, such as (a) low cost and
ready availability of reagents, (b) atom economy, and (c) high
yields (approaching 100%), are applied to rationalize the
(polar or radical) mechanism and to develop new greener
synthetic routes for the synthesis of substituted nitroaromatic
organophosphorus compounds. The nucleophiles used to
phane, triethylphosphane, triphenylphosphane, diphenyl-
phosphane, trimethyl phosphite, triethyl phosphite, dimethyl
phosphonate, diethyl phosphonate, oxo(diphenyl)phos-
phorane, with two nitroaromatic compounds 1,3,5-trinitro-
benzene and 1-chloro-2,4,6-trinitrobenzene in a DMF solu-
tion containing 0.1
M tetrabutylammonium tetrafluoroborate.
In all cases, in order to establish the feasibility or benefits of
study the feasibility and viability of the reaction are the clas- the electrochemical approach relative to the chemical ap-
sical tervalent phosphorus nucleophiles: trimethylphos-
proach, blank reactions were also performed.
In order to develop a general and environmentally
friendly route based on the SNAr reaction, we have focused
our attention on the nucleophilic aromatic substitution re-
action of either hydrogen (NASH process) or a heteroatom
(NASX process) by using electrochemical methods
(Scheme 1).[5–14] In previous studies, we described electro-
chemical cyanation, amination, formation of ketones, and
the synthesis of alkylnitroaromatic compounds. Therefore,
Introduction
The nucleophilic aromatic substitution reaction is one of
the most thoroughly studied reactions, because of its indus-
trial and academic importance.[1] There are numerous syn-
theses of novel compounds and natural products for phar-
maceutical or medical applications, polymers, and analyti-
cal standards for environmental studies based on this classi-
cal reaction.[2] However, the main drawbacks related to the
nucleophilic aromatic substitution reaction are the lack of
generality, the low yield, and the use of external chemical
oxidants, which are hazardous in most cases.[3] Most of the
studies devoted to the selection of the best conditions (sol-
vent, nucleophiles and reactants) and procedures for envi-
ronmental amelioration require mechanistic investigations
to establish the factors that affect the choice of mechanistic
path as well as the regioselectivity of the reaction.[3] It has
been well established by using common spectroscopic tech-
niques such as UV/Vis, IR, and NMR spectroscopy[4] that
when the aromatic ring is substituted with two or more ni-
tro (NO2) groups, the nucleophilic aromatic substitution re-
action involves σ complexes (or Meisenheimer complexes).
The most common mechanism, the SNAr addition-elimi-
nation mechanism, involves two steps. In the first step, there
is a nucleophilic attack on the aromatic ring. In the second
step, the Ar–X bond breaks, whereby the aromaticity of the
nitroarene is recovered.
[a] Departament de Química, Universitat Autònoma de Barcelona,
08193 Bellaterra, Barcelona, Spain
Fax: +34-935812920
E-mail: Iluminada.Gallardo@uab.es
Gonzalo.Guirado@uab.es
Scheme 1.
7378
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 7378–7389

Electrochemical Synthesis of Organophosphorus Compounds through SNAr
the electrochemical approach to SNAr seems to be a highly
When neutral nucleophiles (e.g. amines) are used as nu-
attractive approach from a mechanistic and synthetic point cleophiles, an additional step (prior to the formation of σ
of view. However, the nucleophilic aromatic substitution re- complexes or Meisenheimer complex) occurs, and a new in-
action of hydrogen (NASH process) or heteroatom (NASX termediate is formed.[1–4] These intermediates are com-
process) with nucleophiles containing phosphorus, either monly known as zwitterionic complexes (ZW) (Scheme 3).
chemically or electrochemically, is, as far as we are aware, In most cases, this type of intermediate has been tentatively
limited to a very small number of mechanistic kinetic[15,18] characterized by spectroscopic techniques.[1–4] The same in-
and synthetic studies.[19–22]
termediates are also proposed when imines[23] and phos-
In this regard, in order to obtain esters of substituted phanes[15] are used. However, it is only possible to fully
phosphonic acids, which are used in many chemical and characterize these complexes when they are stable in their
medical applications (for instance, as an agent in the treat- solid form.[23]
ment of osteoporosis),[19] the vicarious nucleophilic aro-
When phosphane acts as a nucleophile, the low stability
matic substitution approach is used (Scheme 2).[20–21] In of the intermediates formed, the low extent of the nucleo-
these NASH reactions, a C–C bond is created instead of philic attack, and the production of undesired products de-
a C–P one when dialkylbenzyl phosphonates are used as rived from secondary reactions are the main drawbacks as-
nucleophiles in the presence of external bases and at low sociated to the use of the SNAr reaction to obtain organo-
temperatures. The σH complex, formed by addition of the phosphorus compounds by direct conversion of nitro com-
carbanion of the benzyl phosphonates to nitroarene, can pounds in nitro compounds containing phosphorus
evolve into the final product with use of a second equivalent (Scheme 4A). In the study of the reaction of P(OAlk)3 with
of base to promote the β-elimination of HX (vicarious nu- 1,3,5-trinitrobenzene in DMSO, the authors proposed the
cleophilic substitution, VNS, Scheme 2A)[21] or by using formation of picryl phosphonate NASH products, although
chemical oxidants, such as potassium permanganate (oxi- they did not consider the mechanism of transforma-
dative nucleophilic substitution of hydrogen, ONSH, tion.[15,18]
Scheme 2B).[22]
When
1-fluoro-2,4,6-trinitrobenzene
and
P(OMe)3 were used, the products obtained were picryl
phosphonate, 2,4,6-trinitrotoluene, and 1-fluoro-2,4,6-trini-
trobenzene. The formation of the picryl phosphonate deriv-
ative as well as MeF can be rationalized in terms of an
SNAr reaction through an Arbuzov rearrangement
(Scheme 4B).[15,18] There are also some studies based on the
reaction between 1,3,5-trinitrobenzene and dialkyl phos-
phonates, (AlkO)2P(O)H. This reaction is promoted in the
presence of base and undergoes conversion to picryl phos-
phonate compounds in DMSO (Scheme 4C).[15,18] The na-
ture of the products formed after either NASH or NASX
shown in Scheme 4 as well as the oxidizing character of the
nucleophile seem to indicate that the SNAr reaction may
involve some radical mechanism running at the same time
as the classic polar mechanism. In this regard, the use of
electrochemical techniques may be an alternative method
or approach for obtaining nitroaromatic organophosphorus
compounds.
Furthermore, the electrochemical anodic oxidation of
phosphorus nucleophiles leads to the formation of radical
cations as a result of the removal of an electron from the
unshared electron pair of the phosphorus atom. After that,
the radical cation can abstract a hydrogen atom from the
Scheme 2.
Scheme 3.
Eur. J. Org. Chem. 2011, 7378–7389
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
7379

H. Cruz, I. Gallardo, G. Guirado
FULL PAPER
Scheme 4.
Scheme 5.
7380
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 7378–7389

Electrochemical Synthesis of Organophosphorus Compounds through SNAr
solvent and finally evolve to the formation of the corre- Table 1. Anodic peak potential, cathodic peak potential, standard
sponding oxidized derivative (Scheme 5).[24]
So, the aim of this study is to develop a synthetic electro-
chemical route to obtain organophosphorus compounds
potential, ΔEp, and number of electrons involved in each electron
transfer of phosphorus compounds and nitroaromatic compounds
in DMF and 0.1 m TBABF4 at 20 °C (scan rate 1.0 Vs–1).
Compound
1
2
3
4
5
6
through the NASH or NASX process following either polar
or radical mechanisms. The presentation of the interesting Epa (V) vs. SCE
1.15
10
1.00
11
1.28
1.36
1.41
1.26
possibilities of the SNAr reaction would require a previous
electrochemical study of the reactants (phosphane com-
pounds used as nucleophiles) and nitroaromatic com-
pounds (electron-deficient aromatic compounds), interme-
diates (σH complexes and zwitterionic intermediates), and
NASH and NASX products (nitroaromatic organophos-
phorus compounds). The mechanistic details and the syn-
thetic scope of the electrochemical method are studied for
a wide series of phosphorus nucleophiles 1–9: trimeth-
ylphosphane (1), triethylphosphane (2), triphenylphos-
phane (3), diphenylphosphane (4), trimethyl phosphite (5),
triethyl phosphite (6), dimethyl phosphonate (7), diethyl
phosphonate (8) and oxo(diphenyl)phosphorane (9) with
two nitroaromatic compounds, 1,3,5-trinitrobenzene (10)
and1-chloro-2,4,6-trinitrobenzene (11), in a DMF solution
containing 0.1 m tetrabutylammonium tetrafluoroborate
(TBABF4). In all cases, in order to establish the feasibility
or benefits of the electrochemical approach relative to the
chemical approach, blank reactions were also performed.
Compound
12
13
1st
wave
1st
wave
1st and 2nd
waves
1st and 2nd
waves
Epc (V) vs. SCE
E° (V) vs. SCE[a]
–0.56 –0.59
–0.37
–0.88 –0.41 –0.93
[a] E° = (Epc + Epa)/2.
triethyl phosphite (6). Controlled-potential electrolysis after
the oxidation peak potential of trialkyl-, triphenyl-, and tri-
alkylphosphanes and phosphites increases the acidity of the
solvent and leads to the corresponding oxides after the
chemical treatment of the electrolyzed sample, as previously
mentioned in the Introduction section (Scheme 5). In those
cases, for nucleophiles 1–6, the oxidation potential values
obtained for each phosphane, which are depicted in Table 1,
would help to determine the excess nucleophile present in
the reaction mixtures.
Results and Discussion
In the current study, cyclic voltammetry was used in a
first stage as an analytical tool, in such a way as to reveal
the concentration of the different species present in the re-
action mixture (reactants: nitroaromatic compounds and
nucleophiles), σH complexes, and zwitterionic intermedi-
ates. For this purpose, it is mandatory to previously study
the electrochemical features of the reactants, products, and
intermediates.
Electrochemical Behavior of Phosphorus Nucleophiles 1–9
The same general trend is observed when cyclic voltam-
mograms of trialkylphosphanes and phosphites (1–6) are
recorded in DMF at low scan rates. No reduction waves are
observed in the first cathodic scan (until –1.00 V), whereas
an irreversible one-electron oxidation wave appears in the
anodic scan (Table 1). There is a shift in the anodic peak
potential value, Epa, (Table 1) from phosphane 1 to 2. Tri-
ethylphosphane 2 is easier to oxidize, by 150 mV, than 1,
which indicates that the methylene group (–CH2–) causes
an inductive effect that is responsible for the stabilization
of the phosphane cation radical.[25]
Figure 1a shows the electrochemical behavior of tri-
phenylphosphane (3), an irreversible monoelectronic oxi-
dation wave is detected at approximately 1.28 V vs. SCE. In
the case of compound 4, where a phenyl group is replaced
with hydrogen, there is a shift to more positive values. The
same general trend is also observed for trialkyl phosphites,
Figure 1. Cyclic voltammograms of 3 (a) showing the oxidation, 11
(b) showing the reduction and 12 (c) showing the reduction in
DMF and 0.1 m TBABF4 on a glassy carbon disk (0.5 mm dia-
trimethyl phosphite (5) being more difficult to oxidize than meter); scan rate 0.50 Vs–1.
Eur. J. Org. Chem. 2011, 7378–7389
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
7381

H. Cruz, I. Gallardo, G. Guirado
FULL PAPER
However, it is remarkable that the electrochemical study diates that will be involved in the nucleophilic aromatic sub-
in terms of the cyclic voltammetry of dimethyl phosphonate stitution reaction: σH complexes and zwitterionic com-
(7), diethyl phosphonate (8), and oxo(diphenyl)phos- plexes.
phorane (9) enables us to conclude that there is no elec-
troactive group up to 1.5 V vs. SCE, since no oxidation
waves are detected in this scan range.
Electrochemical Behavior of Phosphorus σH Complexes
The phosphorus σH complex depicted in Table 2, Entry
1 is carefully prepared by adding two equivalents of potas-
sium tert-butoxide to a mixture of DMF and 0.1 m
TBABF4 solution that contains two equivalents of phos-
phorus nucleophile 7 and one equivalent of nitroaromatic
compound 10, under an argon atmosphere. Analysis of the
cyclic voltammogram of the mixture enabled us to deter-
mine the percentage of 10 remaining in solution, that is, the
extent of the nucleophilic attack (by determining the cur-
rent value of the cathodic peak of 10 when an initial cath-
odic scan is recorded), the type(s) of substitution prod-
uct(s), that is, trinitrocyclohexadienyl σH complexes or σX
complexes, for which Epa is expected to be between 0.5 and
1.7 V,[5,10] and the type(s) of product(s) obtained after the
oxidation of the complex [NASH or NASX product (which
in this case would be reactant 10) can be identified by grow-
ing a reduction wave in a second cycle after the oxidation
of the σH complex or after thorough controlled-potential
electrolysis of the sample]. Figure 2 shows the absence of
electroactive species 10 in the first cathodic cycle, which
means that the percentage of the nucleophilic attack is
100%. Moreover, in the corresponding anodic counterscan,
there are two oxidation peaks; an irreversible two-electron
Electrochemical Behavior of Nitroaromatic Reactants and
Products
The electrochemical behavior of 1,3,5-trinitrobenzene
(10) has been previously published.[26] An irreversible one-
electron wave appears at –0.56 V vs. SCE (Table 1). The oxi-
dation wave observed in the anodic scan at 0.26 V vs. SCE
corresponds to the oxidation of a biradical bis(1,3,5-trini-
trobenzene) dianion. This biradical was characterized as a
π-dimer, which was obtained by the dimerization of two
trinitrobenzene radical anion units. In the case of 1-chloro-
2,4,6-trinitrobenzene (11), an irreversible one-electron wave
is also obtained at –0.59 V vs. SCE (Figure 1b, Table 1).
However, the oxidation wave observed for 11 corresponds
to the oxidation of the chlorine ion.[13] It is important to
remark that for compounds 10 and 11, no oxidation wave
is detected when the cyclic voltammetry begins after an an-
odic scan.
Compound 12, diethyl(2,4,6-trinitrophenyl) phos-
phonate, is chosen as a model compound to characterize
the electrochemical behavior of phosphorylated nitroaro-
matic derivatives. This organophosphorus compound shows
two successive reversible one-electron waves at –0.37 and
–0.88 V vs. SCE (Figure 1c, Table 1). The fact that the re-
duction potential of the expected substitution product is
considerably lower than that of the nitroaromatic reactant
as well as the reversibility of the wave would make it pos-
sible to easily distinguish whether the nucleophilic aromatic
substitution reaction has been completed or not. It is also
important to highlight that no oxidation waves are observed
for any of the phosphorylated nitroaromatic derivatives 12,
13, and 14 in an initial anodic scan. Therefore, none of the
nitroaromatic reagents (10–11) and products (12–14) are ox-
idizing at this potential range.
Having established the electrochemical oxidation features
of the phosphorus nucleophiles and the electrochemical re-
duction and oxidation behavior of nitroarenes, we now
Figure 2. Cyclic voltammograms of a mixture of 10 (1 equiv.), 7
(2 equiv.), potassium tert-butoxide (2 equiv.) in DMF, and 0.1 m
TBABF4 on a glassy carbon disk (0.5 mm diameter). Scan rate
study the electrochemical features of the different interme- 1.0 Vs–1. Scan range 0.0/–1.0/0.5/0.0 V (2 cycles).
Scheme 6.
7382
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 7378–7389

Electrochemical Synthesis of Organophosphorus Compounds through SNAr
oxidation peak at 0.88 V vs. SCE and a small oxidation pre- first reversible reduction wave and the third wave at –0.42
peak at 0.65 V vs. SCE. The cyclic voltammogram recorded and –0.88 V vs. SCE, respectively, could be assigned to the
in a second cycle indicates the formation of a double-substi- phosphorus NASH product 13, whereas the second
tution product, since there are three reduction waves. The irreversible reduction peak at –0.56 V vs. SCE should corre-
Table 2. Products and yields after exhaustive electrolysis of SNAr intermediates.
Eur. J. Org. Chem. 2011, 7378–7389
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
7383

H. Cruz, I. Gallardo, G. Guirado
FULL PAPER
Table 2. (Continued)
[a] The σ complexes were carefully prepared by addition of the nucleophile to nitroarenes, the mixture was dissolved in DMF/0.1 m
TBABF4 under an inert atmosphere at 10 °C. [b] The main electroactive intermediates present in the mixture. [c] The oxidation products
were analyzed by cyclic voltammetry, gas chromatography (GC), 1H and 31P NMR spectroscopy and compared with pure samples
available in the laboratory. [d] Shift of the equilibrium to the right. [e] Blank reactions (without oxidation of the mixture) led to depicted
yields of substitution products. All the potentials are given vs. the SCE reference electrode.
7384
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 7378–7389

Electrochemical Synthesis of Organophosphorus Compounds through SNAr
spond to 10 (Figure 2, second cycle and Table 1). Analysis after controlled-potential electrolysis of the same reactants
of the electrolyzed sample at 1.00 V vs. SCE after the pass- (Table 2, Entry 4). It is remarkable that, despite the low
age of 2F and chemical treatment of the sample (see Experi- yield of NASH product obtained, this procedure is the
mental Section) reveals that two products are obtained: 13 cleanest reaction described in the literature for the synthesis
(80%, NASH product) and 10 (Scheme 6). The fact that the of 14, since reactant 10 is recovered at the end of the pro-
nucleophilic attack is 100% and there is a 20% yield of 10 cess. It can therefore be concluded that the use of bulky
may be due to the formation of an undesired σ complex. arylphosphanes reduces the percentage of nucleophilic at-
The oxidation of this complex, which leads to compound tack from the phosphorus center as well as the yield of
10, may be related with the substitution product formed phosphorylated nitroaromatic derivatives, mainly because
after the oxidation of the first anodic peak at 0.65 V vs. of steric effects.
SCE. When the number of equivalents of base is reduced
from 2 to 1, the percentage of nucleophilic attack decreases
to 80%, the yield of the electrochemically promoted NASH
Electrochemical Behavior of Zwitterionic Compounds with
Trialkyl-, Triaryl- and Diarylphosphanes as Nucleophiles
reaction being 70% (Table 2, Entry 2).
The zwitterionic compounds are carefully prepared “in
situ” by adding different amounts of trialkylphosphanes (1
and 2) to 10. Figure 3 (solid line) shows the cyclic voltam-
mogram of a 1:1 mixture of 1 and 10 in DMF. From the
cathodic scan it can be deduced that the nucleophilic attack
is nonquantitative (55%), since the reduction peak –0.56 V
corresponds to the reduction of 10. The first oxidation peak
observed in the anodic counterscan at 0.26 V is also related
with the electrochemical reactivity of the anion radical of
10 [oxidation of biradical bis(1,3,5-trinitrobenzene) di-
anion]. The second oxidation wave at 1.15 V vs. SCE corre-
sponds to the concentration of 1, which remains “free” in
solution. Thus, the oxidation peak at approximately 1.4 V
as well as the reduction peak at –0.8 V correspond to the
oxidation and reduction peak potential values of the zwit-
terionic compound (Scheme 7).[23] The addition of one
more equivalent of 1 increases the nucleophilic attack to
90% as well as the concentration of zwitterionic compound
(Figure 3, dotted line).
Electrochemically Promoted Nucleophilic Aromatic
Substitution of Hydrogen (NASH Processes) from σH
Complexes and Zwitterionic Complexes with Phosphorus
Nucleophiles in the Presence or Absence of an External
Base – Synthetic Scope
This section reports for the first time in terms of cyclic
voltammetry and preparative electrolysis on the scope and
limitations of electrochemical NASH with phosphorus nu-
cleophiles.
Electrochemical Synthesis of Diethyl(2,4,6-trinitrophenyl)
Phosphonate Oxide (12) and Diphenyl(2,4,6-
trinitrophenyl)phosphane Oxide (14)
The electrosynthesis of diethyl(2,4,6-trinitrophenyl)
phosphonate (12) (in DMF containing 0.1 m TBABF4 with
use of 10 as a nitroarene, 8 as a nucleophile, and potassium
tert-butoxide as a base at a ratio of 1:2:2) is performed by
following the procedure described in the previous section
for obtaining compound 13. The cyclic voltammogram of
the mixture reveals that the percentage of nucleophilic at-
tack is 81% and there are of two oxidation peaks at approx-
imately 0.66 and 0.87 V. This second oxidation peak should
correspond to the phosphorus σH complex (Table 2, Entry
3). After the cyclic voltammetry analysis, the sample is fully
oxidized by thorough controlled-potential electrolysis at
1.00 V vs. SCE. Finally, after chemical treatment and analy-
sis of the sample by cyclic voltammetry, H and 31P NMR
1
Figure 3. Cyclic voltammograms of a mixture of 10 (1 equiv.) and
1 (1 equiv.) in DMF and 0.1 m TBABF4 on a glassy carbon disk
(0.5 mm diameter); scan rate 1.0 Vs–1. Solid line: 0.0/–0.7/1.5/0.0 V.
Dotted line: after the addition of one more equivalent of 1; scan
range 0.0/–0.7/1.5/0.0 V.
spectroscopy enabled confirmation of the synthesis of 12 in
33% yield. It is therefore possible to conclude from the re-
action yields of Entries 1–3 that the yield of a NASH reac-
tion is lower when a diethyl phosphonate is used instead of
a dimethyl phosphonate, probably due to steric effects.
An exhaustive controlled-potential electrolysis at approx-
In order not only to assess the importance of the steric imately 1.50 V leads to the reactant, 10, in 100% yield.
effects but also to extend the methodology to diarylphos- Then, the number of equivalents of 1 is increased to 10, 20,
phane oxide derivatives, we decided to evaluate the use of 30, and 40 equiv. so as to shift the equilibrium from the
oxo(diphenyl)phosphorane (9) as a nucleophile instead of zwitterionic complex to the σ complexes. However, in no
dialkyl phosphonates 7 and 8. In this case, the cyclic vol- case is the substitution product obtained. It is important to
tammogram of the mixture (a DMF solution containing remark that the exhaustive oxidation of the sample corre-
0.1 m of TBABF4, 10 as a nitroarene, 9 as a nucleophile, sponds to the oxidation of the zwitterionic intermediate and
and potassium tert-butoxide as a base at a ratio of 1:2:2) to the excess of nucleophile, so reactant 10 is fully recovered
reveals that, although the percentage of nucleophilic attack at the end of the process and the solvent becomes more
is 80%, only 10% of the NASH product 14 is obtained acidic. This increase in acidity could be explained by the
Eur. J. Org. Chem. 2011, 7378–7389
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
7385

H. Cruz, I. Gallardo, G. Guirado
FULL PAPER
Scheme 7.
oxidation of phosphane leading to the corresponding oxide
and a proton (Scheme 5). The same results were obtained
when the trialkylphosphanes were replaced with 3 and 4
and ratios of nitroaromatic 10 to phosphanes from 10 to 40
were used. In all cases, compound 10 is recovered in 100%
yield and no substitution product is detected (Table 2, En-
tries 5–8).
After disclosing that the electrochemical oxidation of the
σ complexes yields the substitution product by formally re-
placing a hydride (H–) following a NASH mechanism and
that the electrochemical oxidation of the zwitterionic com-
pounds leads to the initial nitroaromatic compound, the ev-
ident extension of this work is to check whether it is pos-
sible to obtain substituted products by means of electro-
chemical oxidation methods, in SNAr reactions with trialkyl
phosphite nucleophiles.
Figure 4. Cyclic voltammograms of a mixture of 10 (1 equiv.) and
6 (0.25 equiv.) in DMF and 0.1 m TBABF4 on a glassy carbon disk
(0.5 mm diameter); scan rate 1.0 Vs–1, scan range 0.0/–1.0/0.5/
0.00 V (2 cycles).
Compound 12 was also synthesized by using the same
methodology as described previously for the synthesis of
13. The reaction yield was 20% when 50 equiv. of 5 were
Electrochemical Synthesis of Dialkyl(2,4,6-trinitrophenyl)
Phosphonates 12 and 13
Figure 4a shows the cyclic voltammetry of a reaction used (Table 2, Entry 12). The analysis by cyclic voltamme-
mixture of 10 (10 mm) in DMF containing 0.1 m of try before performing the electrolysis enabled us to calculate
TBABF4 after the addition of 6 (0.25 equiv.) and prior to the extent of the nucleophilic attack, to characterize the σH
the electrochemical oxidation of the solution. In a first cath- complex and the NASH product 12. Moreover, the cyclic
odic scan, the excess unreacted nitroaromatic compound 10 voltammogram recorded after the passage of 1F confirms
is observed (reduction peak at –0.56 V). Moreover, the for- the oxidation of the zwitterionic complex, the formation of
mation of the zwitterionic complex can be deduced from the NASH product 12 and the recovery of unreacted 10.
the peak current value of the oxidation wave at 0.88 V vs. Blank experiments were also performed to test whether
SCE. The fact that the oxidation potential value is higher there is any NASH process in the absence of the electro-
than that associated to the σH complex, as well as previous chemical oxidation, in no case is the yield of 13 or 12 higher
1H NMR and UV/Vis spectroscopic studies performed in than 3% (Table 2, Entries 13, 14, and 15).
aprotic solvents, led us to conclude that, in this case, the
The proposed oxidation mechanism to electrochemically
intermediate complex detected is a zwitterionic complex synthesize dialkyl(2,4,6-trinitrophenyl) phosphonates in-
rather than a σH complex.[15] The oxidation peak at 0.26 V volves an initial nucleophilic attack of the phosphorus nu-
vs. SCE is associated to the electrochemical oxidation of cleophile. Since there is no detection of the more easily oxi-
the π-dimer dianion, as previously mentioned. In a second dizable corresponding complex by cyclic voltammetry when
cathodic scan, the new first reduction wave, presumably cor- an excess of nucleophile is used, it is reasonable to consider
responding to the substitution product 13, is observed at that there is 100% zwitterionic complex. Finally, the elec-
approximately –0.47 V vs. SCE. Finally, controlled-poten- trochemical oxidation mechanism, which leads to the
tial electrolysis at 1.00 V vs. SCE/1F, after the zwitterionic NASH products, should lead to a radical intermediate after
intermediate oxidation wave, and analysis of the sample (af- removal of an electron and loss of a proton. The displace-
ter a chemical treatment) makes it possible to obtain substi- ment of a methoxy radical from the phosphorus atom of a
tution product 13 in low yield (20%, Table 2, Entry 12). In phosphoranyl radical can easily occur by an α-scission of
order to increase the reaction yield, excess nucleophile is the MeO–P bond. The fragmentation is a regiospecific pro-
added to the DMF solution of 10. The addition of 25 equiv. cess[27] and will finally evolve into the corresponding NASH
of 6 increases the NASH product yield up to 35% (Table 2, products 12 and 13 (Scheme 8) following further oxidation
Entry 10), whereas the addition of 200 equiv. leads to a processes and a similar mechanism to that previously de-
68% yield (Table 2, Entry 11).
scribed for the phosphane cations (Scheme 5).
7386
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 7378–7389

Electrochemical Synthesis of Organophosphorus Compounds through SNAr
Scheme 8.
Electrochemically Promoted Nucleophilic Aromatic
Substitution of Heteroatom (NASX Processes) from
Zwitterionic Compounds with Phosphorus Nucleophiles –
Mechanistic Investigations and Synthetic Scope
5 equiv. of 6 is the same as described previously for Fig-
ure 5. The presence of substitution product 13, unreacted
starting material 11, excess nucleophile 6, and the zwitter-
ionic complex formed can be distinguished.
The electrochemical NASX reaction can be a good alter-
native for surpassing the yields obtained when the electro-
chemical NASH approach is used, although this approach
is less attractive in terms of designing an environmentally
friendly route, since the leaving group is different from hy-
drogen. This situation will be illustrated by synthesizing
compounds 12, 13, and 14.
Electrochemical Synthesis of Dialkyl(2,4,6-trinitrophenyl)
Phosphonates 12 and 13 and Diphenyl-
(2,4,6-trinitrophenyl)phosphane Oxide (14)
Figure 5. Cyclic voltammograms of a mixture of 11 (1 equiv.) and
5 (5 equiv.) in DMF and 0.1 m TBABF4 on a glassy carbon disk
(0.5 mm diameter); scan rate 1.0 Vs–1, scan range 0.0/–1.0/1.5/0.0 V.
Solid line: before electrolysis, dotted line: after electrolysis at 1.50 V
vs. SCE.
By using trialkyl phosphites (5 and 6) as nucleophiles
and 1-chloro-2,4,6-trinitrobenzenes (11) as nitroaromatic
starting materials, it is possible to obtain the NASX substi-
tution products (12 and 13), by either a chemical or an elec-
trochemical route. The chemical reaction of 11 with
50 equiv. of 5 yields 85% 13 and 15% 10 in DMF contain-
ing 0.1 m TBABF4, following an SNAr reaction through an
Arbuzov rearrangement (Table 2, Entry 15). However, the
electrochemical approach makes it possible to reach quanti-
tative yield of compound 13 and even reduce the numbers
of equivalents of 5 (Table 2, Entry 16). Figure 5 shows a
cyclic voltammogram of a mixture containing compound
11 and 5 equiv. of 5. Starting from a cathodic scan, the
presence of two reversible reduction peaks can be distin-
guished at –0.48 and –0.95 V vs. SCE, which correspond
to the formation of the radical anion and dianion of the
substitution product 13, respectively. Moreover, it is also
possible to detect a peak at –0.55 V vs. SCE, which is attrib-
uted to the unreacted starting material, 11. In the corre-
sponding anodic scan it is possible to distinguish two oxi-
dation peaks at approximately 0.85 and 1.30 V vs. SCE.
These peaks can be assigned to the oxidation of the zwitter-
ionic compound and to the oxidation of the excess 5,
respectively. When triethyl phosphite (6) is used as a nucleo-
phile instead of 5, the same general trends are observed.
Substitution products 12 and 10 are obtained in 85% and
It is important to remark that the cyclic voltammetry
experiments support the hypothesis that substitution prod-
ucts 12 and 13 can be obtained through an Arbuzov re-
arrangement. Since the products obtained are the corre-
sponding dialkyl(2,4,6-trinitrophenyl) phosphonates, PV
compounds, and the electrochemical experiments are per-
formed in the absence of oxygen (argon atmosphere), the
formation of the P=O bond under such experimental condi-
tions confirms that the oxygen comes from the O-alkyl
group. The electrochemical oxidation mechanism should in-
volve the generation of the radical cation of the zwitterionic
compound, which would evolve through an anionic oxi-
dation radical NASX Arbuzov-type rearrangement, yield-
ing the corresponding substitution product after either a
chemical (by the excess trialkyl phosphite) or electrochemi-
cal (at the electrode) (Scheme 9) oxidation step.[28] The ver-
satility of alkyl phosphites and their tolerance to a free radi-
cal methodology makes it possible to achieve homolytic re-
action for synthetic chemistry purposes.
Conclusions
We have established the electrochemical oxidation
15% yield through a chemical process (Table 2, Entry 17), mechanism of nine phosphorus nucleophiles as well as the
whereas the yield increases to 100% when electrochemical electrochemical reduction mechanism of four nitroaromatic
oxidation techniques are used (Table 2, Entry 18). The cy- compounds. We have also disclosed the electrochemical oxi-
clic voltammetry of a mixture containing 1 equiv. of 11 and dation mechanism of phosphorus σH complexes as well as
Eur. J. Org. Chem. 2011, 7378–7389
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
7387

H. Cruz, I. Gallardo, G. Guirado
FULL PAPER
Scheme 9.
zwitterionic intermediates by means of cyclic voltammetry
and controlled-potential electrolysis. It has been shown that
the electrochemical oxidation of σH complexes, formed by
addition of dialkyl phosphonates to basic media, as phos-
phorus nucleophiles, makes it possible to obtain the NASH
products through a two-electron mechanism, which corre-
sponds to the formal elimination of hydride anion (nucleo-
philic aromatic substitution of hydrogen mechanism). In the
case of zwitterionic complexes, the alkyl phosphites have
attacked a substitution position. Those complexes evolve
following a promoted anodic oxidation following nucleo-
philic aromatic substitution of heteroatom, NASX, through
an Arbuzov rearrangement. Finally, a nucleophilic aromatic
substitution of hydrogen mechanism is the operating
mechanism for a zwitterionic complex where the nucleo-
phile has been attached to a non-substitution position.
Hence, the present study not only shows a significant im-
provement in the electrochemical preparation of phosphor-
ylated aromatic substituted compounds but also helps
understand and predict the usefulness or uselessness of
using the nucleophilic aromatic substitution route to obtain
the organophosphorus compound desired. Finally, the cur-
rent approach presents an extension of the electrochemical
Experimental Section
General Remarks
Materials: Anhydrous dimethylformamide (DMF, “Pour syntheses
peptidiques”) and acetonitrile (ACN, “Anhydre pour analyses”)
were purchased from SDS and stored in an inert atmosphere with
molecular
sieves.
Tetrabutylammonium
tetrafluoroborate
(TBABF4) was purchased from Fluka (puris.). These reagents were
used without further purification. 1,3,5-Trinitrobenzene was from
Supelco. Phosphorus nucleophiles (1–9) were purchased from Ald-
rich. All commercially available reactants (1–9, 10) and products
(12–13) were of high purity and were used without further purifica-
tion. We synthesized 1-chloro-2,4,6-trinitrobenzene (11) by follow-
ing the method reported in ref.[12]
Chemical Reactions: The nitroarene (23.5ϫ10–5 mol) was added to
the phosphorus nucleophile (5 equiv.) in acetonitrile (ACN),
purged with nitrogen, and left to react for 2 h in an inert atmo-
sphere whilst stirring. After this time, the nitrogen flow was
stopped, and the reaction was left for 8 more hours to promote the
oxidation of the product formed. Then the reaction was stopped,
and the mixture was subsequently partitioned between water and
dichloromethane. The organic layer was dried with anhydrous so-
dium sulfate, the solvents were evaporated at low pressure, and the
residue was analyzed by 1H NMR and 31P NMR spectroscopy.
This residue was then purified by thin layer chromatography.
methodology for the synthesis of different chemicals that Electrochemical Measurements
can be used in several fields, such as pharmacy, medicine,
or material science.
A conical electrochemical cell equipped with a methanol jacket,
which provides a fixed temperature of 20 °C throughout the experi-
7388
www.eurjoc.org
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2011, 7378–7389

Electrochemical Synthesis of Organophosphorus Compounds through SNAr
ment by means of a thermostat, was used to set up the three-elec-
trode system for cyclic voltammetry. For cyclic voltammetry experi-
ments, the working electrode was a glassy carbon disk of 0.5 mm
diameter. It was polished by using a 1 μm diamond paste. The co-
unterelectrode was a Pt disk of 1 mm diameter. All of the potentials
are measured relative to an aqueous saturated calomel electrode
(SCE) isolated from the working electrode compartment by a salt
bridge. The salt solution of the reference calomel electrode was
separated from the electrochemical solution by a salt-bridge ending
with a frit, made of a ceramic material, which permits ionic con-
ductivity between the two solutions and thus avoids appreciable
contamination. The electrolyte solution present in the bridge (or-
ganic solvent and 0.1 m supporting electrolyte) is the same as that
used for the electrochemical solution so as to minimize junction
potentials. Solutions were prepared in N,N-dimethylformamide
(DMF) as solvent and were purged with nitrogen (or argon) before
each measurement, and nitrogen (or argon) was allowed to flow
over the solution during the measurements.
Diphenyl(2,4,6-trinitrophenyl)phosphane Oxide (14): 1H NMR
(250 MHz, CDCl3): δ = 9.00 (s, 2 H), 7.97 (s, 5 H), 7.76 (s, 5 H)
ppm. 31P NMR (250 MHz, CDCl3): δ = 17.7 ppm.
Acknowledgments
We gratefully acknowledge the financial support of the Ministerio
de Educación y Ciencia of Spain through projects CTQ2006-01040
and CTQ 2009-07469.
[1] J. March, Advanced Organic Chemistry, 4th. Ed., Wiley, New
York, 1992, ch. 13, p. 641.
[2] E. Buncel, J. M. Dust, F. Terrier, Chem. Rev. 1995, 95, 2261.
[3] O. N. Chupakhin, V. N. Charushin, H. C. Van der Plas, Nucleo-
philic Aromatic Substitution of Hydrogen, Academic Press, New
York, 1994.
[4] F. Terrier, Nucleophilic Aromatic Displacement (Ed.: H. Feuer),
VCH, New York, 1991.
Electrosynthesis was carried out in the same cell with a PAR 273A
potentiostat. A graphite rod was used as a working electrode and
Pt as a counter electrode, isolated from the solution by a salt bridge
with electrolyte solution, and the SCE was used as a reference elec-
trode, which was also separated from the solution by a salt bridge.
[5] I. Gallardo, G. Guirado, J. Marquet, Chem. Eur. J. 2001, 7,
1759.
[6] G. Moutiers, J. Pinson, F. Terrier, R. Goumont, Chem. Eur. J.
2001, 7,1712.
[7] I. Gallardo, G. Guirado, J. Marquet, Eur. J. Org. Chem. 2002,
2, 251.
Procedure for the Electrochemical Synthesis of Phosphorus Substitu-
tion Products: A solution of nitroarene (5ϫ10–5 mol) in DMF,
which contained TBABF4 (0.1 m) used as a supporting electrolyte,
was prepared under nitrogen (or argon). The corresponding σH
complex was prepared by careful addition of the corresponding
nucleophile (and potassium tert-butoxide in the indicated cases), at
the desired ratio, to the solution of nitroarene under nitrogen (or
argon). The oxidation peaks of the σH complex or zwitterionic in-
termediates were measured by the cyclic voltammetry technique.
Electrosynthesis of oxidation was then carried out at potentials that
were more positively shifted, by approximately 100 mV, than the
value measured for each σ complex or zwitterionic complex. After
thorough electrolysis of the sample, the mixture was subsequently
partitioned between water and toluene. The organic layer was dried
with anhydrous sodium sulfate, the solvents were evaporated at low
pressure, and the residue was analyzed by CV, GC, 1H and 31P
NMR spectroscopy and compared with pure samples of substitu-
tion products. This residue was finally purified by thin layer
chromatography and the nitroorganophosphorus compounds were
obtained at high purity. The σH complexes and zwitterionic inter-
mediates of NASH processes are stable for more than one hour,
whereas the stability of zwitterionic intermediates of NASX is con-
siderably lower (less than 30 min).
Diethyl(2,4,6-trinitrophenyl) Phosphonate (12): 1H NMR
(250 MHz, CDCl3): δ = 8.66 (d, J = 3.35 Hz, 2 H), 4.33 (m, J =
7.025 Hz, J = 3.025 Hz, 4 H), 1.42 (m, J = 7.025 Hz, J = 8.025 Hz,
6 H) ppm. 31P NMR (250 MHz, CDCl3): δ = –1.80 ppm.
Dimethyl(2,4,6-trinitrophenyl) Phosphonate (13): 1H NMR
(250 MHz, CDCl3): δ = 8.68 (d, J = 3.325 Hz, 2 H), 3.95 (d, J =
11.725 Hz, 6 H). 31P NMR (250 MHz, CDCl3): δ = 1.17 ppm. MS:
m/z(%) = 290 (15)[M – OCH3]+, 275(3), 244(4), 231(100), 199(11),
153 (6), 109 (28).
[8] I. Gallardo, G. Guirado, J. Marquet, Eur. J. Org. Chem. 2002,
2, 261.
[9] H. Cruz, I. Gallardo, G. Guirado, Green Chem. 2011, 13, 2531.
[10] I. Gallardo, G. Guirado, J. Marquet, J. Org. Chem. 2003, 68,
631.
[11] I. Gallardo, G. Guirado, J. Marquet, J. Org. Chem. 2003,
63,7334.
[12] I. Gallardo, G. Guirado, J. Marquet, J. Org. Chem. 2002,
67,2548.
[13] I. Gallardo, G. Guirado, J. Marquet, J. Electroanal. Chem.
2000, 488, 64.
[14] I. Gallardo, G. Guirado, J. Marquet, Eur. J. Org. Chem. 2008,
2463.
[15] G. A. Artamkina, M. P. Egorov, I. P. Beletskaya, Chem. Rev.
1982, 82, 427.
[16] P. P. Onys’ko, Y. G. Gololobov, Z. Obshch. Khim. 1980, 50, 729.
[17] P. P. Onys’ko, Y. G. Gololobov, Z. Obshch. Khim. 1979, 49, 39.
[18] P. P. Onys’ko, Y. G. Gololobov, V. P. Prokopenko, Dokl. Akad.
Nauk SSSR 1977, 237, 105.
[19] H. Fleisch, Breast Cancer Res. 2002, 4, 30.
[20] C. M. Andújar Sánchez, M. J. Iglesias, J. García López, I. J.
Pérez Álvarez, F. López Ortiz, Tetrahedron 2006, 62, 3648.
[21] M. Makosza, K. Wojciechowski, Chem. Rev. 2004, 104, 2631.
[22] M. Makosza, D. Sulikowski, J. Org. Chem. 2009, 74, 3827.
[23] I. Gallardo, G. Guirado, Electrochem. Commun. 2007, 9, 173.
[24] V. A. Zagumennov, E. V. Nikitin, A. S. Romakhin, I. P. Kosa-
chev, Elektrokhimiya 2000, 36, 157.
[25] J. L. Bourdelande, I. Gallardo, G. Guirado, J. Am. Chem. Soc.
2007, 129, 2817.
[26] I. Gallardo, G. Guirado, J. Marquet, N. Vilà, Angew. Chem.
2007, 119, 1343; Angew. Chem. Int. Ed. 2007, 46, 1321.
[27] S. Marque, P. Tordo, Top. Curr. Chem. 2005, 250, 43.
[28] S. Yasui, K. Shioji, M. Tsujimoto, A. Ohno, Tetrahedron Lett.
1996, 37, 1625
Received: September 16, 2011
Published Online: October 25, 2011
Eur. J. Org. Chem. 2011, 7378–7389
© 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
7389