A fast radical chain mechanism in the polyfluoroalkoxylation of aromatics through NO2 group displacement. Mechanistic and theoretical studies

Article abstract of DOI:10.1021/jo048354m

(Chemical Equation Presented) Introduction of polyfluoroalkoxy and polyfluoroalkylthio substituents in aromatic rings can be achieved with mild conditions and short times thorough reaction of concentrated solutions of dinitrobenzenes in DMF with polyfluoro alcohols and polyfluoro thiols in moderate excess, in the presence of excess tetrabutylammonium fluoride as a base. Mechanistic studies suggest that under these conditions a fast radical chain mechanism operates. This mechanism is elicited by oxidation of a Meisenheimer complex and proceeds through a radical aromatic substitution with the polyfluoroalkoxy or the polyfluoroalkylthio radicals as key intermediates. At low concentrations, entrainment can be achieved with superoxide anion. A rationale for this effect is discussed. Answers to particular questions about the proposed mechanism are achieved through a theoretical study at the B3LYP/6-31+G(d,p) level. Specifically, the competition between the radical mechanism and the corresponding polar one (classical S_NAr reaction) is studied in that way, with the conclusion that the key steps of the radical mechanism in our reaction conditions (polar aprotic solvent) are at least as efficient as the ones of the polar one, thus justifying the observed kinetic advantage for the chain reaction in the conditions where an efficient initiation occurs.

Full text of DOI:10.1021/jo048354m

A Fast Radical Chain Mechanism in the Polyfluoroalkoxylation of
Aromatics through NO2 Group Displacement. Mechanistic and
Theoretical Studies
Ismael Tejero, Imma Huertas, AÅ ngels Gonza´lez-Lafont, Jose´ M. Lluch,* and Jordi Marquet*
Departament de Quı´mica, Universitat Auto`noma de Barcelona, 08193 Bellaterra, Barcelona, Spain
lluch@klingon.uab.es; jordi.marquet@uab.es
Received September 16, 2004
Introduction of polyfluoroalkoxy and polyfluoroalkylthio substituents in aromatic rings can be
achieved with mild conditions and short times thorough reaction of concentrated solutions of
dinitrobenzenes in DMF with polyfluoro alcohols and polyfluoro thiols in moderate excess, in the
presence of excess tetrabutylammonium fluoride as a base. Mechanistic studies suggest that under
these conditions a fast radical chain mechanism operates. This mechanism is elicited by oxidation
of a Meisenheimer complex and proceeds through a radical aromatic substitution with the
polyfluoroalkoxy or the polyfluoroalkylthio radicals as key intermediates. At low concentrations,
entrainment can be achieved with superoxide anion. A rationale for this effect is discussed. Answers
to particular questions about the proposed mechanism are achieved through a theoretical study at
the B3LYP/6-31+G(d,p) level. Specifically, the competition between the radical mechanism and
the corresponding polar one (classical SNAr reaction) is studied in that way, with the conclusion
that the key steps of the radical mechanism in our reaction conditions (polar aprotic solvent) are
at least as efficient as the ones of the polar one, thus justifying the observed kinetic advantage for
the chain reaction in the conditions where an efficient initiation occurs.
1. Introduction
stitution reactions between polyfluoroalkoxide anions and
chloro-, fluoro-, or nitro-substituted electron-poor aromat-
ics in dipolar aprotic solvents, mainly HMPA.5,6
Even though nitro group displacement by trifluoro-
ethoxide anion in HMPA is successful at room temper-
ature with a variety of substituted nitrobenzenes, the
reaction is slow, and rather high temperatures (150 °C)
are needed for other polyfluoroalkoxydes, failing the
reaction for branched ones such as hexafluoroisopro-
poxide anion.6d
The nucleophilic displacement of nitro groups in aro-
matic compounds constitutes a simple (in principle) but
fascinating reaction with still many open questions.
Despite the striking difference on nitrite leaving group
ability in aliphatic or aromatic chemistry, it is generally
Considerable time and attention have been and con-
tinue to be directed toward the synthesis of fluorinated
organic materials.1 Lately, in addition to the general
interesting properties of these compounds, the demon-
strated usefulness of biphasic fluorinated chemistry has
added a more interest in their synthesis and the study
of their properties.2 Very recently, the combination of the
advantages of palladium nanoparticles with their recov-
ery in biphasic organic fluorous solvent systems has been
reported.3 In the case of polyfluoroalkoxyaromatics, two
kinds of synthetic approaches have been described: (1)
the reactions of electrophilic haloalkyl fluorides, fluoro-
alkenes, or fluoroalkylsulfonates with nucleophilic phenol
derivatives4,5 and (2) direct aromatic nucleophilic sub-
(1) Organofluorine Chemistry. Principle and Commercial Applica-
tions; Banks, R. E., Smart, B. E., Tatlow, J. C., Eds.; Plenum: New
York, 1994.
(5) Prescher, D.; Thiele, T.; Ruhmann, R. J. Fluor. Chem. 1996, 92,
1619.
(6) (a) Idoux, J. P.; Gupton, J. T.; McCurry, C. K.; Crews, A. D.;
Jurss, C. D.; Colon, C.; Rampi, R. J. Org. Chem. 1983, 48, 3771. (b)
Gupton, J. T.; Idoux, J. P.; DeCrescenzo, G.; Colon, C. Synth. Commun.
1984, 14, 621. (c) Gupton, J. T.; Hertel, G.; Decrescenzo, G.; Colon, C.;
Baran, D.; Dukesherer, D.; Novik, S.; Liotta, D.; Idoux, J. P. Can. J.
Chem. 1985, 63, 3037. (d) Idoux, J. P.; Madenwald, M. L.; Garcia, B.
S.; Chen, D. L.; Gupton, J. T. J. Org. Chem. 1985, 50, 1876.
(2) Horva´th, I. T. Acc. Chem. Res. 1998, 31, 641 and references
therein.
(3) Moreno-Man˜as, M.; Pleixats, R. Acc. Chem. Res. 2003, 36, 638
and references therein.
(4) (a) Rico, I.; Wakselman, C. Tetrahedron Lett. 1981, 22, 323. (b)
Rico, I.; Wakselman, C. Tetrahedron 1981, 37, 4209.
10.1021/jo048354m CCC: $30.25 © 2005 American Chemical Society
Published on Web 02/02/2005
1718
J. Org. Chem. 2005, 70, 1718-1727

Polyfluoroalkoxylation of Aromatics through NO2 Group Displacement
SCHEME 1
accepted that its nucleophilic displacement from electron-
poor aromatics follows the polar SNAr mechanism with
the first step as the rate-limiting step, thus justifying
the just commented difference in behavior.7
In many cases, typical SNAr reactions show an unex-
pected dependence on conditions8,9 and show radical
trends,10-12 radical anions have been detected,13-17 oxygen
has been shown to act as a promoter,18 etc. All this
behavior indicates a mechanistic complexity not yet
resolved, since none of the mechanistic proposals ad-
vanced (generally including the radical anion as inter-
mediate, i.e., the SRN2 chain mechanism,10,19,20 or electron
transfer followed by collapse21) have resisted a close
examination.20,22-24 Among those SNAr reactions with a
peculiar behavior, the nucleophilic displacement of nitro
group is central, with a significant number of examples
described.10,12,16,17,24
TABLE 1. Reactions of Dinitrobenzenes with
Polyfluoro Alcohols and Polyfluorothiols in the Presence
of TBAF‚3H2O in DMFa
reaction
product yield (%) time (h)
entry substrate
nucleophile
1
2
3
4
5
6
7
8
p-DNB CF3CH2OH
1a
1bb
1c
1d
2a
3a
3c
3d
98
1
21
2
p-DNB (CF3)2CHOH
p-DNB CF3(CF2)6CH2OH
p-DNB CF3CH2SH
30 (46)c
89
92
2
o-DNB
CF3CH2OH
76 (86)c
70 (88)c
d
1.5
m-DNB CF3CH2OH
m-DNB CF3(CF2)6CH2OH
m-DNB CF3CH2SH
3
24
5
34 (83)c
a [DNB] ) 0.02 M, [NuH] ) 0.06 M, [TBAF‚3H2O] ) 0.1 M. b In
this reaction, a 11% yield of p-nitrophenol was obtained in addition
to the substitution product. c The yield in parentheses is calculated
with respect to the consumed (nonrecovered) starting material.
d This product was not quantified since it was not possible to purify
it.
We have been interested in the synthetic possibilities
and the mechanistic ambiguities of the nucleophilic
displacement of the nitro group in aromatics for some
time,24,25 and recently, we have described that anodic
oxidation of the σ-complex intermediate in SNAr reactions
produces its instantaneous decomposition leading to
substitution products.25,26
In our experience, and from the results reported in the
literature, the radical trends in SNAr reactions are nor-
mally observed when working with an excess of nucleo-
phile/base. In this work, we describe a synthetic improve-
ment of the known reaction of polyfluoroalkoxylation by
using an excess of nucleophile/base, a change that
transforms the mechanism from a slow polar SNAr to a
very fast chain radical one. An experimental mechanistic
study has been carried out, and the hypothesis advanced
has been analyzed by theoretical methods. The polyfluoro-
alkoxylation reaction, being a relatively slow one, and
the polyfluoroalkoxy anion and radical, being peculiar
intermediates, allow us to observe clearly the change of
mechanism, and the features of the radical one. The
conclusions reached in this paper can probably apply to
other SNAr and related reactions that show narrow
borderlines between polar and radical mechanisms, thus
opening a via to improve our understanding of this
borderline in reactions, such as SNAr, that formally go
through tetrahedral intermediates.
(7) Nucleophilic Aromatic Displacement. The Influence of the Nitro
Group; Terrier, F., Ed.; VCH: New York, 1991.
(8) (a) Shein, S. M.; Brykhovetskaya, L. V.; Pishchugin, F. V.;
Starichenko, V. F.; Panfilov, V. N.; Voevodskii, V. V. J. Struct. Chem.
1970, 11, 228. (b) Blyumenfeld, L. A.; Brykhovetskaya, L. V.; Formin,
G. V.; Shein, S. H. Russ. J. Phys. Chem. 1970, 44, 518.
(9) (a) Iwasaki, G.; Saeki, S.; Hamana, M. Chem. Lett. 1986, 31. (b)
Iwasaki, G.; Wada, K.; Saeki, S.; Hamana, M. Heterocycles 1984, 22,
1811.
(10) (a) Denney, D. B.; Denney, D. Z. Tetrahedron 1991, 47, 6577.
(b) Denney, D. B.; Denney, D. Z.; Perez, A. J. Tetrahedron 1993, 49,
4463.
(11) Zhao, W.; Huang, Z. J. Chem. Soc., Perkin Trans. 2 1991, 1967.
(12) Sammes, P. G.; Thetford, D.; Voyle, M. J. Chem. Soc., Perkin
Trans. 1 1988, 3229.
(13) (a) Bacaloglu, R.; Bunton, C. A.; Cericheli, G. J. Am. Chem.
Soc. 1987, 109, 621. (b) Bacaloglu, R.; Bunton, C. A.; Cericheli, G.;
Ortega, F. J. Am. Chem. Soc. 1988, 110, 3495. (c) Bacaloglu, R.; Bunton,
C. A.; Ortega, F. J. Am. Chem. Soc. 1988, 110, 3503. (d) Bacaloglu, R.;
Bunton, C. A.; Ortega, F. J. Am. Chem. Soc. 1988, 110, 3512.
(14) Bilkis, I. L.; Shein, S. M. Tetrahedron 1975, 31, 969.
(15) Relles, C. F.; Johnson, D. S.; Manello, J. S. J. Am. Chem. Soc.
1977, 99, 6677.
(16) (a) Abe, T.; Ikegami, Y. Bull. Chem. Soc. Jpn. 1976, 49, 3227.
(b) Abe, T.; Ikegami, Y. Bull. Chem. Soc. Jpn. 1978, 51, 196.
(17) Sauer, A.; Wasgestian, F.; Barbasch, B. J. Chem. Soc., Perkin
Trans. 2 1990, 1317.
(18) Omelechko, E. N.; Ryabinin, V. A.; Shein, S. M. Zh. Org. Khim.
1982, 18, 1123.
(19) Marquet, J.; Jiang, Z.; Gallardo, I.; Batlle, A.; Cayo´n, E.
Tetrahedron Lett. 1993, 34, 2801
(20) Galli, C.; Bunnett, J. F. J. Am. Chem. Soc. 1979, 101, 6137.
(21) (a) Zhang, X. M.; Yang, D. L.; Liu, Y. C. J. Org. Chem. 1993,
58, 224. (b) Zhang, X. M.; Yang, D. L.; Liu, Y. C. J. Org. Chem. 1993,
58, 7350.
(22) Makosza, M.; Podraza, R.; Kwast, A. J. Org. Chem. 1994, 59,
6796.
(23) Balslev, H.; Lund, H. Tetrahedron 1994, 50, 7899.
(24) (a) Mir, M.; Espin, M.; Marquet, J.; Gallardo, I.; Tomasi, C.
Tetrahedron Lett. 1994, 35, 9055. (b) Marquet, J.; Casado, F.; Cervera,
M.; Espin, M.; Gallardo, I.; Mir, M.; Niat, M. Pure Appl. Chem. 1995,
67, 703.
(25) Gallardo, I. Guirado, G.; Marquet, J. J. Org. Chem. 2002, 67,
2548.
(26) (a) Gallardo, I.; Guirado, G.; Marquet, J. Chem. Eur. J. 2001,
7, 1759. (b) Huertas, I.; Gallardo, I.; Marquet, J. Tetrahedron Lett.
2001, 42, 3439. (c) Gallardo, I.; Guirado, G.; Marquet, J. Eur. J. Org.
Chem. 2002, 251. (d) Gallardo, I.; Guirado, G.; Marquet, J. Eur. J. Org.
Chem. 2002, 261. (e) Gallardo, I.; Guirado, G.; Marquet, J. J. Org.
Chem. 2003, 68, 631. (f) Gallardo, I.; Guirado, G.; Marquet, J. Org.
Chem. 2003, 68, 7334.
2. Experimental Results and Discussion
2.1. Preparative Reactions. Reactions of m-, o-, and
p-dinitrobenzenes (DNB) with different polyfluoro alco-
hols or polyfluorothiols in the presence of tetrabutyl-
ammonium fluoride (TBAF) as base28 are described in
Scheme 1 and Table 1. A relatively concentrated solution
of DNBs was used (0.02 M), nucleophiles and base were
used in excess with respect to the substrate, and all the
reactions were carried out at room temperature (the
nature of the base has no influence in the reactions since
similar results were achieved using sodium hydride).
Reactions of p-DNB with 2,2,2-trifluoroethanol, 2,2,2-
trifluoroethanethiol, and 1H,1H-pentadecafluoro-1-oc-
tanol are very fast and were completed in less than 2 h
(entries 1, 3, and 4, Table 1). Reaction of p-DNB with a
(27) Schaal, R.; Peure, F. Bull. Soc. Chim. Fr. 1963, 2638.
(28) Clark, J. H. Chem. Rev. 1980, 80, 429.
J. Org. Chem, Vol. 70, No. 5, 2005 1719

Tejero et al.
branched perfluoro alcohol such as 1,1,1,3,3,3-hexafluoro-
2-propanol gave rise to the substitution product in fair
yield, together with minor amounts of p-nitrophenol
(entry 2, Table 1). Notice that, as far as we know, this is
the first example of perfluoroalkoxylation using a
R-branched nucleophile.
As expected, m-DNB was less reactive, but in any case
good preparative yields were obtained with 2,2,2-tri-
fluoroethanol and 2,2,2-trifluoroethanethiol (entries 6
and 8, Table 1). A similar result was obtained in the
reaction of o-DNB and 2,2,2-trifluoroethanol (entry 5,
Table 1).
The reaction of m-DNB with 1H,1H-pentadecafluoro-
1-octanol produced the substitution product in fair yield
(entry 7, Table 1); however, this reaction result could not
be quantified due to the fact that the high viscosity of
the product prevented any proper purification.
FIGURE 1. Evolution of absorbance (λ ) 302 nm, product
1a) vs time (min) in the reaction of p-DNB with 2,2,2-
trifluoroethanol (molar ratio 5:1 with respect to p-DNB), in
the presence of FTBA‚3H2O (molar ratio 5:1 with respect to
p-DNB) in DMF, at different starting substrate concentrations.
These preparative reactions, when compared with
related ones reported in the literature,6 show that the
use of a relatively high concentration of aromatic sub-
strate (0.02 M) and a modest excess of base and nucleo-
phile accelerates significantly the processes, allowing the
reaction to proceed under much milder conditions.
The nature of this faster process was first investigated
by carrying out the reaction of p-DNB and 2,2,2-trifluoro-
ethanol (entry 1, Table 1, 98% yield) in the presence of
variable amounts of galvinoxyl as radical scavenger.
Thus, in the presence of 0.3 equiv (molar ratio with
respect to m-DNB) of galvinoxyl, only a 4% of substitution
product was produced, and upon increasing the amount
of galvinoxyl to 3.0 equiv, the reaction was stopped
completely. Evidently this behavior, similar to other cases
reported in the literature,10-12 suggested the operation
of a radical chain mechanism.
2.2. Mechanistic Studies. There are very few kinetic
studies on nucleophilic aromatic substitutions with ni-
trite anion as a leaving group; however, some of them
are interesting with respect to our own results. Thus,
Schaal and Peure reported good first-order kinetics in
the reaction of p-DNB with methoxide anion in methanol
at concentrations lower than 1 M; however, at higher
concentrations of nucleophile the rate constant increased
exponentially.27 Therefore, we decided to study the effect
of substrate and base concentration on the evolution of
our reactions. No good fit to standard kinetics was
obtained under our conditions.
FIGURE 2. Evolution of absorbance (λ ) 302 nm, product
1a) vs time (min) in the reaction of p-DNB (starting concentra-
tion, 3.0 × 10-4M) with 2,2,2-trifluoroethanol (molar ratio 5:1
with respect to p-DNB), in the presence of variable concentra-
tions of FTBA‚3H2O (molar ratio 5:1, and 10:1, with respect
to p-DNB) in DMF.
ethanol/FTBA is kept at 1:5:5, a relatively slow process
is observed; however, when a relationship of 1:5:10 is
used, the instantaneous process becomes apparent.
The obvious interpretation is the existence of a narrow
mechanistic borderline between a slow mechanism and
a fast, probably radical chain mechanism, considering the
behavior of the preparative reaction in the presence of
the radical scavenger galvinoxyl. This radical scavenger
is not appropriate in reactions to be followed by UV/vis
spectroscopy due to its strong absorption. Therefore, for
qualitative kinetic analysis, sulfur was used. In Figure
In Figure 1, the evolution of the absorbance at λ ) 302
nm (absorption of the substitution product) with time in
the reaction of p-DNB with 2,2,2-trifluoroethanol, at
different starting substrate concentrations (keeping con-
stant the molar relationship p-DNB/trifluoroethanol/
TBAF, 1:5:5), is described. Surprisingly, a sharp change
in behavior is observed when going from 3.0 × 10-4 to
4.0 × 10-4 M. In the first case, a relatively slow reaction
evolution is observed, while upon a small increase in
substrate concentration, the reaction becomes almost
instantaneous.
3 the dependence of the “fast reaction” (5 × 10-4
M
substrate concentration, in the conditions of Figure 1)
on the presence of variable amounts of sulfur is described.
Sulfur is not as strong as galvinoxyl as a radical
scavenger; however, a significant retarding effect can be
observed, even at molar relationships with respect to
substrate, lower than one.
In Figure 2, the evolution of the absorbance at λ ) 302
nm with time in the reaction of p-DNB with 2,2,2-
trifluoroethanol, at different base concentrations, using
a concentration of substrate, p-DNB, 3.0 × 10-4 M (the
fast reaction is not triggered in the conditions used in
the experiments described in Figure 1), is described.
Thus, when the molar relationship p-DNB/trifluoro-
The existence of radical trends in the SNAr reactions
has been in many cases related with the hypothetic
operation of the corresponding radical anions as inter-
mediates. This is very well documented for SRN1 reac-
tions,29 but no consistent confirmation of its general role
has been found in the reactions of nitroaromatics despite
a variety of proposals (see, for instance, refs 7, 8, 10, 13,
1720 J. Org. Chem., Vol. 70, No. 5, 2005

Polyfluoroalkoxylation of Aromatics through NO2 Group Displacement
SCHEME 2
FIGURE 3. Evolution of absorbance (λ ) 302 nm, product
1a) vs time (min) in the reaction of p-DNB with 2,2,2-
trifluoroethanol (molar ratio 5:1 with respect to p-DNB), in
the presence of FTBA‚3H2O (molar ratio 5:1 with respect to
p-DNB) in DMF, with different amounts of added sulfur
(radical scavenger).
Meisenheimer complexes formed on a heteroatom-
occupied position.25,26
(e) Even though in some particular cases nucleophilic
substitution of aromatic halides32 and other nonactivated
aromatics33 has been described to be elicited by oxidative
catalysis, the well-established mechanism for those cases
(SON2)32 includes radical cations of the starting aromatic
compound and of the final substitution product. Such
intermediates have a very low probability to operate in
our reactions since we are dealing with very electron-
poor aromatics (nitroaromatic compounds).
(f) Even though direct oxidation of the nucleophile has
been proposed as initiation step in some cases,34 the
observed concentration effects, the fact that fluoride
anion is not basic enough to form discrete anions of the
nucleophile, and the consideration of the redox potentials
of the nucleophile anion and the electronically excited
(under visible light) Meisenheimer complex make oxida-
tion of the nucleophile in front of oxidation of a Meisen-
heimer complex a low probability initiation step in our
standard reactions (however, see below the example of
entrainment with superoxide anion).
16, 19, and 21). It is known that when a nitroaromatic
compound is placed in front of high concentration of an
anion, the corresponding radical anion can be detected
by ESR spectroscopy.17,30 This was also our case, since
p-DNB in the presence of TBAF under our standard
reaction conditions (1:5 molar relationship), in the ab-
sence of nucleophile, produced the p-DNB radical anion,31
which could be detected by ESR. However, its presence
does not necessarily mean its operation as intermediate.
Indeed, the external electrogeneration of the radical
anion produced an inhibition of the reaction when
compared with a blank. The same happened when the
generation of the radical anion was carried out photo-
chemically with a reducing agent like triethylamine
(Scheme 1). This behavior confirms similar results de-
scribed in related systems.23,24
To propose a mechanistic hypothesis able to explain
the experimental evidence, several points must be taken
into account.
In Scheme 2, a mechanistic hypothesis for the initiation
steps of the mechanism is proposed. The Meisenheimer
complex derived from p-DNB and fluoride anion would
undergo oxidation by p-DNB,30 and the fast evolution25
of the resulting Meisenheimer radical would give rise to
the nitrogen dioxide radical that would enter in the
propagation steps of the mechanism. Even though the
second step in Scheme 2 is probably endergonic, the fact
that Meisenheimer complexes absorb strongly in the
visible region (the reactions are carried out under ambi-
ent light) will make this electron-transfer step thermo-
dynamically favorable. In addition, and in any case, an
efficient radical chain may considerably amplify a weak
initiation.35
(a) Fluoride anion is a very hard anion, and its
probability to participate in outer sphere electron-
transfer processes must be negligible. On the other hand,
it is a good nucleophile in polar aprotic solvents.28
(b) The double concentration effect, with respect to the
substrate and with respect to the base, suggest a radical
chain mechanism with an initiation step that is very
dependent on both reagents.
(c) Scorrano’s interpretation for the appearance of
radical anions in the reaction mixtures of SNAr processes
of nitroaromatics includes a bimolecular oxidation of
the Meisenheimer complex by its own starting nitro-
aromatic.30 In our case, this could explain the observed
concentration effects.
In Scheme 3, the proposed propagation steps of the
radical chain are described. Indeed, the mechanism
(d) Some of us have reported that stable Meisenheimer
anion complexes, derived from nitroaromatic compounds,
upon anodic oxidation undergo a fast decomposition,
giving rise to substitution products. This decomposition
takes place via elimination of a radical in the case of
(32) (a) Alder, R. W. J. Chem. Soc., Chem. Commun. 1980, 1184.
(b) Eberson, L.; Jo¨nsson, L. J. Chem. Soc., Chem. Commun. 1980, 1187.
(c) Eberson, L.; Jo¨nsson, L. J. Chem. Soc., Chem. Commun. 1981, 133.
(d) Eberson, L.; Jo¨nsson, L.; Wistrand, L. G. Tetrahedron 1982, 38,
1087. (e) Jo¨nsson, L.; Wistrand, L. G. J. Org. Chem. 1984, 49, 3340.
(33) den Heijer, J.; Shadid, O. B.; Cornelisse, J.; Havinga, E.
Tetrahedron 1977, 33, 779.
(34) Russell, G. A. J. Am. Chem. Soc. 1954, 76, 1595.
(35) Costentin, C.; Hapiot, P.; Me´debielle, M.; Save´ant, J. M. J. Am.
Chem. Soc. 1999, 121, 4451.
(29) (a) Bunnett, J. F. Acc. Chem. Res. 1978, 11, 413. (b) Save´ant,
J. M. Tetrahedron 1994, 50, 10117.
(30) Mariani, C.; Modena, G.; Pizzo, G. P.; Scorrano, G. J. Chem.
Soc., Perkin Trans. 2 1979, 1187.
(31) High-Resolution ESR Spectroscopy; Gerson, F., Ed.; Wiley: New
York, 1970.
J. Org. Chem, Vol. 70, No. 5, 2005 1721

Tejero et al.
SCHEME 3
we have theoretically studied both the anionic and the
radical addition-elimination mechanisms which are
initiated by nucleophilic or radical attack at the π-aro-
matic molecule p-DNB of, respectively, the alkoxide
anions or the alkoxy radicals derived from ethanol and
2,2,2-trifluoroethanol. In this section, we will first provide
the computational details and then the theoretical re-
sults.
3.1. Computational Details. The quantum mechan-
ical calculations have been carried out within the frame-
work of the density functional theory (DFT),40 which is
reaching a widespread use in the calculation of quite
sizable organic and inorganic molecules. The spin-
unrestricted formalism has been used in solving the
Kohn-Sham DFT equations.41 The particular functional
used has been the Becke three-parameter hybrid method
using the Lee, Yang, and Parr correlation functional
(B3LYP),42 nowadays one of the most used functionals.
The basis set chosen for the calculations had to be flexible
enough to describe anionic species and, therefore, the
split valence 6-31+G (d,p) basis set,43 which includes d
and p polarization functions on heavy and hydrogen
atoms, respectively, and a diffuse sp shell on the heavy
atoms, has been used. The diffuse functions provide
ample space allowance for the additional electron in the
anions. A preliminary Hartree-Fock study using the
same basis set has also been carried out for the reactions
of ethoxide anion and ethoxy radical.
corresponds to a typical radical aromatic substitution,36
although from the theoretical calculations described
below it is not clear if the Meisenheimer radical operates
as an intermediate or as a transition state. A downhill
(see the Theoretical Calculations section) outer sphere
electron transfer between an anion and a radical initiates
the cycle.
It is known that superoxide anion is a good oxidant
for relatively acidic protic substrates to the corresponding
radical through a process that includes dismutation.37,38
Therefore, trifluoroethanol radical could be generated
alternatively by the action of superoxide anion on tri-
fluoroethanol (pKa ) 23.6 in DMSO).39 The fast reaction
between p-DNB and trifluoroethanol could be elicited by
addition of a catalytic amount of KO2 (and an equi-
molecular amount of the 18-crown-6 ether to achieve its
solubilization in DMF). This represents a good example
of entrainment, and considering superoxide anion has no
outher sphere electron-transfer oxidant properties,37 this
result strongly supports the operation of the trifluoro-
ethoxy radical as intermediate.
Two questions appear when considering our mecha-
nistic proposal:
(a) Can the radical aromatic substitution compete with
the polar one?
(b) Is the step that opens the propagation cycle
thermodynamically favorable?
Answers to these two questions have been obtained
through theoretical calculations that are described in the
following section.
Full geometry optimization and direct location of
stationary points (minima and transition state struc-
tures) have been carried out by means of the Schlegel
gradient optimization algorithm by using redundant
internal coordinates.44 Diagonalization of the potential
energy analytical second-derivative matrix (Hessian) has
been done to disclose the nature of the stationary point
of the potential energy surface: no negative eigenvalues
indicate a potential energy minimum, whereas one nega-
tive eigenvalue identifies a transition-state structure. In
this second case, the eigenvector (transition vector)
associated with the negative eigenvalue shows the direc-
tion along which the potential energy lowers. To ensure
the connectivity between each transition state structure
and the corresponding minimum energy points at both
sides of it, the minimum energy path (MEP) has been
calculated in each case by following the Gonza´lez-
Schlegel mass-weighted internal-coordinates reaction-
path algorithm.45 Each MEP has been built up with a
step size of 0.1 bohr amu1/2
.
The thermodynamic magnitudes have been computed
by using the statistical thermodynamic formulation46 of
partition functions within the ideal gas, rigid rotor, and
harmonic oscillator models. A pressure of 1 atm and a
temperature of 298.15 K have been assumed in the
calculations. The B3LYP analytical second derivatives of
3. Theoretical Calculations
The experimental results shown above indicate that
p-DNB reacts with trifluoroethanol in the presence of
base through a radical chain mechanism to give the
corresponding aromatic substitution of the nitro group.
Interestingly enough, related reactions of other alcohols
do not seem to show radical behavior, and the well-known
two-step polar SNAr mechanism has been assigned.7 To
be able to discuss the feasibility of the different mecha-
nisms through which these SNAr reactions can take place,
(40) Kohn, W.; Becke, A. D.; Parr, R. G. J. Phys. Chem. 1996, 100,
12974.
(41) (a) Honenberg, P.; Kohn, W. Phys. Rev. B 1964, 136, 864. (b)
Kohn, W.; Sham, L. J. Phys. Rev. A 1965, 140, 1133.
(42) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(43) Ab Initio Molecular Orbital Theory; Hehre, W. J., Radom, L.,
Schleyer, P. v. R., Pople, J. A., Eds.; Wiley: New York, 1986.
(44) Peng, C.; Ayala, P. Y.; Schlegel, H. B.; Frisch, M. J. J. Comput.
Chem. 1996, 17, 49.
(36) Tiecco, M. Acc. Chem. Res. 1980, 13, 51.
(37) Sawyer, D. T.; Valentine, J. S. Acc. Chem. Res. 1981, 14, 393.
(38) Nanni, E. J., Jr.; Stallings, M. D.; Sawyer, D. T. J. Am. Chem.
Soc. 1980, 102, 4481.
(45) Gonzalez, C.; Schlegel, H. B. J. Phys. Chem. 1990, 94, 5523.
(46) McQuarrie, D. A. Statistical Thermodynamics; University Sci-
ence Books: Mill Valley, CA, 1973.
(39) Bordwell, F. G. Acc. Chem. Res. 1988, 21, 456.
1722 J. Org. Chem., Vol. 70, No. 5, 2005

Polyfluoroalkoxylation of Aromatics through NO2 Group Displacement
the potential energy with respect to the Cartesian
coordinates have been used for the determination of
vibrational frequencies. The imaginary frequency is
neglected in the thermodynamic evaluation for transition-
state structures.
The polarized continuum (overlapping spheres) model
(PCM)47 has been used to study the solvent effect for most
of the reaction systems. Both electrostatic and nonelec-
trostatic terms have been included in the calculation.
Acetonitrile (ꢀ ) 36.64), a solvent of the list included in
the Gaussian 98 package48 whose dielectric constant
value is very close to the DMF one, has been employed
as solvent. As a matter of fact, several experiments were
also done in acetonitrile, leading to results similar to the
ones obtained using DMF as solvent. For the first
propagation step of the radical chain, the self-consistent
isodensity PCM (SCI-PCM)49 model, whose solvent cavity
is determined self-consistently from an isodensity surface,
FIGURE 4. HF/6-31+G(d,p) gas-phase potential energy
has been used.
diagrams for the anionic (ethoxide anion) and radical (ethoxy
radical) processes. Reactants RA and RR are taken as origin of
energy for the anionic and the radical reactions, respectively.
All values are given in kcal/mol.
The Gaussian 98 package48 has been used to carry out
all of the electronic calculations.
3.2. Theoretical Results. To begin, we will focus on
the reaction of p-DNB with ethoxide anion and with
ethoxy radical in gas phase. For the sake of comparison,
we will show in parallel the results corresponding to the
anionic and radical reactions. We have located 7 and 6
stationary points for the two processes at the Hartree-
Fock electronic level, respectively, which leads to the
potential energy profiles shown in Figure 4. From here
on the subscripts A or R will mark the structures located
along the anionic or the radical mechanisms, respectively.
In the first case, the pathway evolves through an ion-
molecule complex in the entrance channel (reactant
complex, RCA), a first transition-state structure (TS1A),
an intermediate (IntA, which is a Meisenheimer σ-com-
plex), a second transition- state structure (TS2A), and an
ion-molecule complex in the exit channel (product
complex, PCA), giving finally the final products (PA, that
is, p-ethoxynitrobenzene and nitrite anion). All these
structures are much lower in potential energy than the
reactants (RA), as the global process is very exoergic. This
is due to the fact that the net negative charge becomes
more delocalized in going from reactants to products.
The sequence of stationary points along the radical
pathway is analogous (but the reactant complex RCR does
not exist as a minimum), although the relative potential
energies respect to p-DNB and ethoxy radical (RR) are
at least about 40 kcal/mol higher than the corresponding
relative potential energies with respect to p-DNB and
ethoxide anion along the anionic pathway. This is a
consequence of the absence of a net negative charge that
migrates during the process. Anyway, the global radical
reaction turns out to be also exoergic leading to p-
ethoxynitrobenzene and nitrite radical. Note that the
neutral intermediate IntR could be formally considered
as an oxidated Meisenheimer complex (it has lost an
electron). On the other hand, we have to point out that
these Hartree-Fock calculations give a huge spin con-
tamination of the radical species (expected values of
within the range 1.2-1.5, instead of 0.75 which corre-
sponds to a pure doublet) which indicates that a higher
electronic level of calculation is strongly required for
treating these processes in a suitable way.
The above-described Hartree-Fock picture clearly
changes when electron correlation is introduced by means
of B3LYP calculations (see Figure 5 for potential energies
and Figures 6 and 7 for the geometries of the located
stationary points).
As for the anionic mechanism, due to the lack of TS2A,
the Meisenheimer σ-complex is not longer a minimum.
Then, only a transition-state structure (TSA) has been
located in a very flat region, with a geometry similar to
the TS1A previously found at the Hartree-Fock electronic
level. It has to be emphasized that correlation energy
turns out to be essential in order to discuss the nature
of the Meisenheimer complex. It is a shallow minimum
at the Hartree-Fock electronic level, but it simply
disappears as stationary point (neither an intermediate
nor a transition-state structure) at the B3LYP electronic
level. Likewise, the oxidated Meisenheimer complex is
not a stationary point (due to the absence of TS2R) for
the B3LYP radical pathway too, for which only a transi-
tion-state structure (TSR) appears after a reactant com-
plex (RCR) which did not exist at the Hartree-Fock
electronic level. Analogously to the anionic case, the
geometry of TSR looks like that of TS1R, but it is different
from the geometry of the IntR found at the Hartree-Fock
electronic level.
(47) (a) Miertus, S.; Scrocco, E.; Tomasi, J. Chem. Phys. 1981, 55,
117. (b) Miertus, S.; Tomasi, J. Chem. Phys. 1982, 65, 239.
(48) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A.,
Jr.; Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.;
Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.;
Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo,
C.; Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.;
Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;
Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; Johnson, B. G.; Chen,
W.; Wong, M. W.; Andres, J. L.; Head-Gordon, M.; Replogle, E. S.;
Pople, J. A. Gaussian 98, revision A.7-A.11; Gaussian, Inc.: Pitts-
burgh, PA, 1998.
From the energetic point of view, although the par-
ticular figures change, the anionic pathway is still
(49) Foresman, J. B.; Keith, T. A.; Wiberg, K. B.; Snoonian, J.;
Frisch, M. J. J. Phys. Chem. 1996, 100, 16098.
J. Org. Chem, Vol. 70, No. 5, 2005 1723

Tejero et al.
FIGURE 5. B3LYP/6-31+G(d,p) gas-phase potential energy
diagrams for the anionic (ethoxide anion) and radical (ethoxy
radical) processes. Reactants RA and RR are taken as origin of
energy for the anionic and the radical reactions, respectively.
All values are given in kcal/mol.
FIGURE 7. B3LYP/6-31+G(d,p) stationary-point structures
along the radical (ethoxy radical) reaction path. All distance
values given are in Å.
FIGURE 8. B3LYP/6-31+G(d,p)-solvated (ꢀ ) 36.64) energy
diagrams for the anionic (ethoxide anion) and radical (ethoxy
radical) processes. Reactants RA and RR are taken as origin of
energy for the anionic and the radical reactions, respectively.
All values are given in kcal/mol.
FIGURE 6. B3LYP/6-31+G(d,p) stationary point structures
along the anionic (ethoxide anion) reaction path. All distance
values given are in Å.
notoriously more favorable than the radical one in terms
of relative potential energy at the B3LYP electronic level.
In all cases the B3LYP spin contamination along the
whole radical pathway was negligible ( S2 values never
exceeded 0.751).
Until now we have only presented gas-phase calcula-
tions. To assess the solvent effect (acetonitrile), we have
solvated the B3LYP gas-phase stationary points, their
geometries being kept frozen. The new energy profiles
are shown in Figure 8.
The energies of the reactant complexes (RCA and RCR)
become higher than their respective reactants (RA and
RR), in such a way that they are no longer minima and
they have not been represented in Figure 8. Likewise,
the product complex PCA goes above the product PA and
disappears as a minimum. The effect on the anionic
pathway is quite dramatic as a consequence of the huge
relative stabilization of the species in which the net
negative charge is more localized, that is, nitrite anion
on the product side and, especially, ethoxide anion in the
reactant side. This anionic mechanism already involves
a positive energy barrier, the energy difference between
Solvation destabilizes clearly both the anionic and the
radical pathways with respect to the corresponding
reactants, although the two processes are still exergonic.
1724 J. Org. Chem., Vol. 70, No. 5, 2005

Polyfluoroalkoxylation of Aromatics through NO2 Group Displacement
FIGURE 9. B3LYP/6-31+G(d,p) gas-phase potential energy
diagrams for the anionic (2,2,2-trifluoroethoxide anion) and
radical (2,2,2-trifluoroethoxy radical) processes. Reactants RA
and RR are taken as origin of energy for the anionic and the
radical reactions, respectively. All values are given in kcal/
mol.
FIGURE 10. B3LYP/6-31+G(d,p) stationary-point structures
along the anionic (2,2,2-trifluoroethoxide anion) reaction path.
All distance values are given in Å.
TSR and TSA having been reduced from 43.2 kcal/mol in
gas phase to 10.2 kcal/mol in solution. As a result, the
reaction of p-DNB with ethoxide anion is still predicted
to be ostensibly faster than the reaction with ethoxy
radical, although solvation with a polar solvent has
reduced a lot the energy gap between them in comparison
with the scenario in gas phase.
Let us turn our attention to the reaction of p-DNB with
2,2,2-trifluoroethoxide anion and with 2,2,2-trifluoro-
ethoxy radical. Based on the above-described results, we
have now directly used the B3LYP electronic level to
study these reactions.
The corresponding gas-phase potential energy profiles
and the geometries of the located stationary points are
shown in Figures 9, 10, and 11, respectively. Comparing
Figure 9 with Figure 5, it can be seen that the substitu-
tion of three hydrogen atoms by three fluorine atoms in
the methyl group slightly stabilizes the radical mecha-
nism, but it largely destabilizes the anionic pathway as
a consequence of the electron-withdrawing power of those
three fluorine atoms. As a result of that, the potential
energy difference between TSR and TSA in the gas phase
is now of only 32.7 kcal/mol, 10.5 kcal/mol lower than
for the case of the derivatives of ethanol. In addition, an
inversion in the energetic ordering of the final products
occurs: PA is 16.7 kcal/mol more stable than PR for the
case of ethanol, whereas it becomes 8.7 kcal/mol higher
than PR for the case of 2,2,2-trifluoroethanol.
The sequence and nature of the stationary points is
the same as in the reactions with the derivatives of
ethanol. Comparing geometries, TSA for the reaction of
2,2,2-trifluoroethanolate anion (the distances for the O-C
forming and the C-N breaking bonds are, respectively,
1.947 and 1.501 Å) appears to be somewhat displaced
toward products relative to the same transition state
structure for the reaction of ethanolate anion (the
distances for the O-C forming and the C-N breaking
bonds are, respectively, 2.158 Å and 1.468 Å). Conversely,
FIGURE 11. B3LYP/6-31+G(d,p) stationary-point structures
along the radical (2,2,2-trifluoroethoxy radical) reaction path.
All distance values are given in Å.
TSR for the reaction of 2,2,2-trifluoroethoxy radical seems
to be very slightly moved toward reactants respect to TSR
for the reaction of ethoxy radical. These facts are in
agreement with the Hammond’s postulate.
When the B3LYP gas-phase stationary points are
solvated, the effects for the reactions with the derivatives
of 2,2,2-trifluoroethanol (the energy profiles are pictured
J. Org. Chem, Vol. 70, No. 5, 2005 1725

Tejero et al.
4. Conclusions
We have demonstrated that polyfluoroalkoxylation of
aromatics can be carried out by mild substitution of a
nitro group in nitroaromatic compounds in DMF, under
conditions that include a relatively high concentration
of substrate, nucleophile, and base. Mechanistic studies
indicate the operation of a fast radical chain mechanism
that includes an oxidation step; therefore, it seems we
are in the presence of an example of reaction that is
activated by oxidation (“one electron less”).50 Due to its
biological relevance, the effect of superoxide anion as an
activator is particularly significant.
The proposal for the propagation part of the mecha-
nism has been studied by theoretical calculations, and
the conclusion is that the radical substitution step is at
least competitive with the corresponding reaction of the
trifluoroethoxide anion. Therefore, it becomes evident
that the chain mechanism can have a kinetic advantage
once the proper initiation conditions are reached. The
mechanism proposed, or a related one, can probably
explain many experimental facts reported in the litera-
ture that suggest the existence of a radical mechanism
in SNAr reactions in certain conditions.
The theoretical study presented here is also a nice
example of how the successive incorporation of different
factors can even revert the conclusion first reached when
using the lowest level of calculation, and conditions (gas
phase), far away from the experimental reality (polar
aprotic solvent).
FIGURE 12. B3LYP/6-31+G(d,p)-solvated (ꢀ ) 36.64) energy
diagrams for the anionic (2,2,2-trifluoroethoxide anion) and
radical (2,2,2-trifluoroethoxy radical) processes. Reactants RA
and RR are taken as origin of energy for the anionic and the
radical reactions, respectively. All values are given in kcal/
mol.
in Figure 12) are quantitatively and qualitatively quite
similar to the case of ethanol.
Due to solvation complexes RCA, RCR and PCA disap-
pear and both pathways are destabilized relative to the
corresponding reactants, the effect on the anionic mech-
anism being again quite striking. This time the energy
difference in the gas phase (32.7 kcal/mol) between TSR
and TSA almost vanishes, becoming as small as 1.5 kcal/
mol in acetonitrile as a solvent. Finally, we have calcu-
lated the zero-point energies, the thermal corrections, and
the entropic contributions at the gas-phase stationary
points in order to evaluate the Gibbs free energy barriers.
When those three terms are added to the values given
in Figure 12 we show that, in terms of free energy for
the solvated reactions at 298.15 K, the anionic pathway
is exergonic (∆G° ) -14.6 kcal/mol) with a Gibbs free
energy barrier of 23.8 kcal/mol, whereas the radical
mechanism is also exergonic (∆G° ) -21.3 kcal/mol) with
a Gibbs free energy barrier of 23.6 kcal/mol. Very
interestingly, we realize at this point that both barriers
are equal in practice, although strictly speaking they
have even reversed the ordering that had been found up
to now in this paper.
We have also calculated the reaction energy of the
formation of the 2,2,2-trifluoroethoxy radical according
to the first propagation step of the radical chain (see
Scheme 3). A B3LYP/6-31+G(d,p) calculation inside a
solvent cavity (with full geometry optimization) leads to
a reaction energy (without zero-point energy and thermal
corrections) of -3.43 kcal/mol. This value confirms the
feasibility of this propagation step as source of the 2,2,2-
trifluoroethoxy radicals required to produce the radical
substitution reaction.
5. Experimental Section
Commercial o-, m-, and p-dinitrobenzenes, 2,2,2-trifluoro-
ethanol, 1,1,1,3,3,3-hexafluoro2-propanol, 1H,1H-pentadeca-
fluoro-1-octanol, 2,2,2-trifluoroethanthiol, and tetrabutyl-
ammonium fluoride (as trihydrate, TBAF‚3H2O) were used
without further purification.
General Procedure for the Reactions of Dinitroben-
zenes with Polyfluoro Alcohols and Polyfluoro Thiols
Described in Table 1. In a 100 mL round-bottomed flask
with magnetic stirring was introducted 5.0 mmol of TBAF‚
3H2O in 45 mL of dry DMF. Then, 3.0 mmol of the nucleophile
and 1.0 mmol of the DNB in 5 mL of dry DMF were added.
The mixture was left under magnetic stirring for the time
indicated in Table 1. Then, the reaction mixture was acidified
with aqueous HCl and was extracted several times with ether.
The organic layer was extracted again with aqueous HCl and
dried and the solvent was evaporated, leaving the crude
product that was purified from the remaining starting material
(if any) and from the 4-nitrophenol (if any) by pressure column
chromatography through silica gel using mixtures of hexane/
ethyl acetate of increasing polarity as eluent.
p-(2,2,2-Trifluoroethoxy)nitrobenzene, 1a: mp 75.5-
1
76.5 °C (lit.6a mp 75-78 °C); H NMR (CDCl3) δ 4.43 (q, J )
8.0 Hz, 2H), 7.02 (d, J ) 8.8 Hz, 2H), 8.23 (d, J ) 8.8 Hz, 2H);
MS m/z (relative intensity) 221 (M+, 100), 191 (41), 156 (10),
111 (20), 92 (34), 83 (24), 64 (29).
Finally, it has to be noted that our B3LYP results
predict that both the anionic and the radical pathways
are concerted, with the bond formation and bond cleavage
happening in the same step. So, the second and third
propagation steps in Scheme 3 could be reduced to an
unique step. In other words, it seems that, in this case,
the reaction system crosses very fast through the region
of the potential energy surface corresponding to the
oxidated Meisenheimer structure, which is not a stable
complex.
p-(2,2,2-Trifluoro-1-trifluoromethylethoxy)nitroben-
zene, 1b: viscous oil; IR (KBr) 1594, 1525, 1494, 1350, 1257,
1200, 1175, 886, 800, 663 cm-1; 1H NMR (CDCl3) δ 5.00 (septet,
J ) 2.7 Hz, 1H), 7.18 (d, J ) 9.3 Hz, 2H), 8.26 (d, J ) 9.3 Hz,
2H); 13C NMR (CDCl3) δ 74.6, 116.9, 120.7, 126.2, 144.3, 161.2.
Anal. Calcd for C9H5F6NO2: C, 39.58; H, 1.85; N, 5.13.
Found: C, 39.91; H, 1.98; N, 5.32.
(50) Chanon, M.; Rajzman, M.; Chanon, F. Tetrahedron 1990, 46,
6193.
1726 J. Org. Chem., Vol. 70, No. 5, 2005

Polyfluoroalkoxylation of Aromatics through NO2 Group Displacement
p-(2,2,3,3,4,4,5,5,6,6,7,7,8,8,8-Pentadecafluorooctyloxy)-
nitrobenzene, 1c: mp 49.5-50.0 °C (lit.51 mp 42-45 °C); 1H
NMR (CDCl3) δ 4.55 (t, J ) 12.4 Hz, 2H), 7.03 (d, J ) 9.5 Hz,
2H), 8.23 (d, J ) 9.5 Hz, 2H).
(the structure assignation must be considered as tentative);
1H NMR (CDCl3) δ 4.54 (t, J ) 12.5 Hz, 2H), 7.29 (ddd, J )
1.0, 2.5, 8.4 Hz, 1H), 7.50 (dd, J ) 8.0, 8.4 Hz, 1H), 7.78 (dd,
J ) 2.0, 2.5 Hz, 1H), 7.93 (ddd, J ) 1.0, 2.0, 2.0, 1H); MS m/z
(relative intensity) 521 (M+, 100).
p-(2,2,2-Trifluoroethylthio)nitrobenzene, 1d: viscous
oil; IR (KBr) 1597, 1518, 1342, 1313, 1241, 1133, 1083, 845,
m-(2,2,2-Trifluoroethylthio)nitrobenzene, 3d: viscous
1
742, 639 cm-1; H NMR (CDCl3) δ 3.58 (q, J ) 9.5 Hz, 2H),
oil; IR (KBr) 1528, 1350, 1264, 1241, 1125, 1081, 802, 731 cm-1
;
7.49 (d, J ) 9.5 Hz, 2H), 8.16 (d, J ) 9.5 Hz, 2H); 13C NMR
(CDCl3) δ 37.7, 124.2, 124.9, 128.7, 143.1, 146.5. Anal. Calcd
for C8H6F3NO2S: C, 40.51; H, 2.55; N, 5.91. Found: C, 40.89;
H, 2.73; N, 6.15.
1H NMR (CDCl3) δ 3.52 (q, J ) 9.4 Hz, 2H), 7.51 (t, J ) 7.8
Hz, 1H), 7.76 (m, 1H), 8.13 (m, 1H), 8.30 (t, J ) 1.8 Hz, 1H);
13C NMR (CDCl3) δ 37.3, 122.6, 124.2, 125.0, 125.4, 130.05,
136.2, 136.6. Anal. Calcd for C8H6F3NO2S: C, 40.51; H, 2.55;
N, 5.91. Found: C, 40.37; H, 2.66; N, 5.68.
o-(2,2,2-Trifluoroethoxy)nitrobenzene, 2a. This product
was isolated by column chromatography as indicated in the
“general procedure” and purified by distillation (114 °C, 3
Torr):6a 1H NMR (CDCl3) δ 4.47 (q, J ) 8.0 Hz, 2H), 7.14 (m,
2H), 7.57 (dt, J ) 7.3, 2.2 Hz, 1H), 7.85 (dd, J ) 8.0, 2.2 Hz,
1H).
Acknowledgment. We are grateful for financial
support from the Spanish “Ministerio de Ciencia y
Tecnolog´ıa” and the “Fondo Europeo de Desarrollo
Regional” through Project Nos. BQU2002-00301 and
BQU2003-05457 and from the “Generalitat de Cata-
lunya” through Project No. 2001SGR 00180. Likewise,
we acknowledge the use of the computational facilities
of the CESCA.
m-(2,2,2-Trifluoroethoxy)nitrobenzene, 3a: mp 58.0-
1
58.5 °C (lit.6a mp 57-59 °C); H NMR (CDCl3) δ 4.43 (q, J )
8.0 Hz, 2H), 7.29 (dd, J ) 8.0 Hz, J ) 2.2 Hz, 1H), 7.49 (t, J
) 8.0 Hz, 1H), 7.77 (t, J ) 2.2 Hz, 1H), 7.93 (dd, J ) 8.0 Hz,
J ) 2.2 Hz, 1H); MS m/z (relative intensity) 221 (M+, 100),
175 (32), 111 (31), 92 (35), 83 (16).
Supporting Information Available: Tables 1-36 show
the geometries, energies, and number of imaginary vibrational
frequencies of the stationary points located. This material is
available free of charge via the Internet at http://pubs.acs.org.
m-(2,2,3,3,4,4,5,5,6,6,7,7,8,8,8-Pentadecafluorooctyl-
oxy)nitrobenzene, 3c: viscous oil that could not be purified
(51) Prescher, D.; Thiele, T.; Ruhmann, R. J. Fluorine Chem. 1996,
79, 145.
JO048354M
J. Org. Chem, Vol. 70, No. 5, 2005 1727