Heterogeneous photoinduced homolytic aromatic substitution of electron-rich arenes with perfluoroalkyl groups in water and aqueous media - A radical-ion reaction

Article abstract of DOI:10.1002/ejoc.201201271

The photoinduced electron transfer (PET) substitution reaction of electron-rich aromatic nuclei with perfluoroalkyl R_f groups was carried out in water or aqueous mixtures to render substitution products resulting from replacement of aromatic H atoms with the R_f moiety in good yields (57-88 %). Some mechanistic aspects are discussed, supporting the notion of a PET reaction leading to a classical radical homolytic aromatic substitution (HAS) followed by an electron transfer (ET) and then a proton transfer (PT) sequence. A radical mechanism superimposed on a redox process is proposed to account for product formation. Evidence for the radical cation species (as an initiation event) generated from electron-rich arenes in the presence of perfluoroalkyl halides is provided by the UV/Vis transient spectra obtained by Nanosecond Laser Flash Photolysis techniques. Copyright

Full text of DOI:10.1002/ejoc.201201271

Job/Unit: O21271
/KAP1
Date: 18-12-12 18:15:16
Pages: 12
FULL PAPER
DOI: 10.1002/ejoc.201201271
Heterogeneous Photoinduced Homolytic Aromatic Substitution of Electron-
Rich Arenes with Perfluoroalkyl Groups in Water and Aqueous Media – A
Radical-Ion Reaction
Sebastián Barata-Vallejo,[a] Marina Martín Flesia,[a] Beatriz Lantaño,[a] Juan E. Argüello,[b]
Alicia B. Peñéñory,*[b] and Al Postigo*[a]
Keywords: Heterogeneous catalysis / Green chemistry / Radical reactions / Radical ions / Electron transfer / Alkylation
The photoinduced electron transfer (PET) substitution reac-
tion of electron-rich aromatic nuclei with perfluoroalkyl Rf
groups was carried out in water or aqueous mixtures to
render substitution products resulting from replacement of
aromatic H atoms with the Rf moiety in good yields (57–
88%). Some mechanistic aspects are discussed, supporting
the notion of a PET reaction leading to a classical radical
homolytic aromatic substitution (HAS) followed by an elec-
tron transfer (ET) and then a proton transfer (PT) sequence.
A radical mechanism superimposed on a redox process is
proposed to account for product formation. Evidence for the
radical cation species (as an initiation event) generated from
electron-rich arenes in the presence of perfluoroalkyl halides
is provided by the UV/Vis transient spectra obtained by
Nanosecond Laser Flash Photolysis techniques.
Introduction
Studer and Curran[1] have recently argued in favor of the
homolytic aromatic substitution (HAS) pathway of many
reported non-metal-mediated radical C–H activation or or-
ganocatalytic reactions leading to arylation of aromatic
rings. It is well-accepted that the key step in these HAS
reactions is the addition of an aryl radical to an arene, re-
sulting in cyclohexadienyl-type radicals that promptly re-
aromatize. The possible mechanisms for rearomatization
are well-known and long-established in the literature, i.e.,
disproportionation, reaction with an initiator by hydrogen
transfer, and oxidation.[2] Examples of the former three re-
aromatization processes of the cyclohexadienyl-type radi-
cals in water or aqueous mixtures are documented (see be-
low).
Rearomatization of the intermediate cyclohexadienyl
radical can occur by deprotonation to render the radical
anion of the substitution product, followed by electron
transfer (ET), or through ET to afford the cyclohexadienyl
cation (Wheland intermediate), followed by proton transfer
(PT), as shown in Scheme 1.[1]
Scheme 1. PT and ET of cyclohexadienyl-type radicals.
In contrast to radical perfluoroalkylation addition reac-
tions of unsaturated compounds,[3,4a] fluoroalkyl radical
substitution on the aromatic ring (HASs with perfluoro-
alkyl groups) has received little attention.[4b] Furthermore,
aromatic radical perfluoroalkylation in aqueous mixtures is
restricted to a few examples of sulfinatodehalogenation re-
actions.[4b,5]
Radical perfluoroalkylation reactions involving unsatu-
rated compounds in aqueous media have also been re-
viewed.[6] In particular, activated aromatic nuclei, such as
1,3,5-trimethoxybenzene, were employed in perfluoroalkyl
substitution radical reactions in aqueous media. When a
mixture of 1,3,5-trimethoxybenzene and 1-bromo-1-chloro-
2,2,2-trifluoroethane was treated under sulfinato-dehalo-
genation reaction conditions (Na2S2O4/NaHCO3 in MeCN/
H2O), a radical substitution reaction of H by Rf was
achieved.[7] The authors[7] postulate a radical HAS mecha-
nism such as that depicted in Scheme 2.
[a] Departamento de Química Orgánica, Facultad de Farmacia y
Bioquímica, Universidad de Buenos Aires,
Junín 954 CP 1113, Buenos Aires, Argentina
Fax: +54-11-4964-8250
E-mail: apostigo@ffyb.uba.ar
Homepage: www.uba.edu.ar
[b] INFIQC, Departamento de Química Orgánica, Facultad de
Ciencias Químicas, Universidad Nacional de Córdoba, Ciudad
Universitaria,
5000 Córdoba, Argentina
Homepage: www.unc.edu.ar
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201201271.
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1

Job/Unit: O21271
/KAP1
Date: 18-12-12 18:15:16
Pages: 12
A. B. Peñéñory, A. Postigo et al.
FULL PAPER
In this account, we will demonstrate that for electron-
rich nuclei such as N,N-dialkyl-substituted aromatic amines,
N-substituted carbazoles, and methoxy-substituted aro-
matics, PET does take place in water or aqueous media in
the presence of perfluoroalkyl iodides (and bromides), ren-
dering this methodology appropriate for the synthesis of
fluorinated aromatics in the heterogeneous water environ-
ment. Evidence for a radical-ion process and singlet excited
state reactivity will be provided.
Scheme 2. Mechanism for the aromatic radical substitution reac-
tion of H for Rf.[7]
Results and Discussion
Aromatic Substitution Reactions of N,N-Dialkyl-Substituted
Aromatic Amines, N-Substituted Carbazoles, and Methoxy-
Substituted Aromatics with RfX in Water and Aqueous
Media
In another report by the same leading author,[8] the aro-
matic substitution of trimethoxybenzenes, and pyrrols with
BrCF2CF2Br in aqueous media was attempted rendering
the mono-Rf-substituted trimethoxybenzene and the 2-sub-
stituted pyrrols in 99 and 85% yields, respectively.
When an Ar-degassed heterogeneous mixture of N,N-di-
methylaniline 1 or N,N-dimethyl-1-naphthylamine 2 (2–
5 mmol) and n-C4F9I (1 mmol) in water (30 mL) was vigor-
ously stirred in a photoinduced reaction [unfiltered me-
dium-pressure Hg lamp (MPL),[13b] 4 h, 20 °C], the nona-
fluorobutyl para substitution product 3 and the 4-substitu-
tion product 4, respectively, were obtained in yields ranging
from 57 to 88%, based on n-C4F9I (Scheme 3 and Table 1,
entries 1 and 2).
More recently, Lü and co-workers[9] have accomplished
the polyfluoroalkylation of 2-aminothiazoles and deriva-
tives. Reactions of 2-aminothiazole with n-C4F9I under
sulfinatodehalogenation reaction conditions, afforded the
substitution product in 80% yield with total selectivity at
the 5-position of 2-aminothiazole. A number of N-substi-
tuted 2-aminothiazoles also react with RfI and RfBr in
yields ranging from 58 to 90%.[9]
Chen and co-workers[10] achieved the radical perfluoro-
alkylation reaction of aromatic amines by substitution of
aromatic H atoms for Rf groups in dimethyl sulfoxide
(DMSO) as solvent, triggered through the decomposition
of Na2S2O4. They attained a series of 2-substituted 1,4-di-
aminobenzenes with Rf moieties that, upon further reac-
tion, afforded intramolecular cyclization products.[11] In all
these examples, the rearomatization of the cyclohexadienyl
radical intermediate is achieved by hydrogen transfer.
More recently, Yamakawa and co-workers have ac-
complished the fluorination of aromatic compounds using
the Fenton reagent in DMSO as solvent, in a metal-induced
electron-transfer-like reaction.[12a,12b] They attempted a
series of simple-substituted aromatic and heteroaromatic
compounds with CF3I and ethoxycarbonyl difluoromethyl
bromide as fluorinating agents. In this case, the rearomati-
Scheme 3. Perfluoroalkyl group substitution of aromatic amines in
water.
The product yields increased upon increasing the concen-
zation of the cyclohexadienyl radical intermediate is tration of substrates, and the best product yields were ob-
achieved by metal-induced oxidation and then PT.
tained when the substrate/n-C4F9I ratio was ca. 5:1.
We also undertook the photoreaction of 2 and n-C4F9I
A photochemical method was successfully employed to
effect perfluoroalkylation reactions of aromatic and het- in Ar-deoxygenated water with filtered (350 nm) fluorescent
eroaromatic compounds to provide substitutions with the lamps (2ϫ 40 W), and obtained a similar substitution yield
corresponding α-aryl-α,α-difluoroacetates and aryl-α,α-di- of 4 (Table 1, entry 3), demonstrating that the lower emis-
fluoromethylphosphonates in good to moderate yields in sion wavelengths of the medium pressure Hg lamp (i.e.,
CH2Cl2 as solvent.[12c] Photonucleophilic aromatic substitu- 313 nm) did not influence the product yield or distribution
tion of 6-fluoroquinolones has also been successfully ac- (see Exp. Sect. for lamp details).
complished in water.[13a]
To cast some light on the reactive excited state manifold
To the best of our knowledge, metal-free photoinduced of substrate 1 (or 2), we carried out product studies of the
electron transfer (PET) methodology has not been em- PET reaction (at irradiation wavelength λ = 310 nm) of 1
ployed in water or aqueous mixtures to accomplish the per- (1 mmol), in the presence of n-C4F9I (1 mmol) and 4-meth-
fluoroalkyl substitution reaction of aromatic nuclei ef- oxyacetophenone (MAP; 15 mmol), which is a well-known
ficiently.
triplet energy sensitizer (ET = 310 kJmol–1). Under these
2
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O21271
/KAP1
Date: 18-12-12 18:15:16
Pages: 12
Aromatic Substitution of Arenes with Perfluoroalkyl Groups
Table 1. Reaction conditions tested in the HAS reaction of arenes with perfluoroalkyl halides in heterogeneous media under vigorous
stirring.
[b]
[c]
Entry
Synthetic method
Substrates [mmol]
Solvent system [mL]
Product yield (%)[a] Asubstrate:ARfX pHinitial/pHfinal
1
2
3
4
5
6
7
8
MPL[d]
1 (5), n-C4F9I (1)
2 (5), n-C4F9I (1)
2 (5), n-C4F9I (1)
H2O (30)
H2O (30)
H2O (30)
3 (88)
4 (57)
4 (63)
3 (–)
5.5:2.5[c]
MPL[d]
5.5:2.5[c]
λ = 350 nm[e]
λ = 310 nm
λ = 310 nm
MPL[d]
5.5:2.5[c]
1 (1), n-C4F9I (1), MAP[f] (15) H2O (30)
5.5:5.5[c]
2 (1), n-C4F9I (1), MAP[f] (15) H2O (30)
4 (–)
5.5:5.5[c]
1 (5), n-C8F17I (1)
1 (5), n-C8F17Br (1)
2 (5), n-C8F17I (1)
2 (5), n-C8F17Br (1)
1 (0.1), n-C4F9I (10)
2 (0.1), n-C4F9I (10)
1 (2.5), n-C4F9I (0.5)
2 (2.5), n-C4F9I (0.5)
1 (1), n-C4F9I (0.33)
2 (1), n-C4F9I (0.33)
1 (1), n-C4F9I (0.33)
2 (1), n-C4F9I (0.33)
1 (5), n-C4F9I (1)
2 (5), n-C4F9I (1)
7 (1), n-C4F9I (0.25)
9 (5), n-C4F9I (1)
10 (5), n-C4F9I (1)
13 (5), n-C4F9I (1)
14 (5), n-C4F9I (1)
15 (2.5), n-C4F9I (0.5)
H2O (30)
H2O (30)
H2O (30)
H2O (30)
H2O (4)
H2O (4)
H2O (4)
H2O (4)
H2O (4)
5 (86)
5 (82)
6 (68)
6 (62)
3 (80)[g]
4 (65)[g]
3 (80)
4 (61)
3 (–)
4 (–)
3 (–)
4 (–)
3 (81)
4 (49)
8 (50)
11 (32)
12 (15)
–
5.5:2.5[c]
MPL[d]
5.5:2.5[c]
MPL[d]
5.5:2.5[c]
9
MPL[d]
5.5:2.5[c]
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
λ = 254 nm
λ = 254 nm
λ = 254 nm
λ = 254 nm
thermal (100 °c)
thermal (100 °c)
ultrasound (80 °c)
ultrasound (80 °c)
MPL[d]
1:10[b]; 5.5:2.5[c]
1:10[b]; 5.5:2.5[c]
100:1[b]; 5.5:2.5[c]
100:1[b]; 5.5:2.5[c]
5.5:5.5[c]
H2O (4)
H2O (4)
H2O (4)
5.5:5.5[c]
5.5:5.5[c]
5.5:5.5[c]
MeCN (30)
MeCN (30)
MeCN/H2O[h] (30)
H2O (30)
H2O (30)
H2O (30)
H2O (30)
H2O (4)
MPL[d]
MPL[d]
5.5:2.5[c]
5.5:2[c]
5.5:2
MPL[d]
MPL[d]
MPL[d]
5.5:2[c]
5.5:2[c]
5.5:2.5[c]
MPL[d]
–
λ = 254 nm
16 (85)[e]
[a] Yield based on n-C4F9I or otherwise indicated. [b] Absorbance ratio of substrate and RfX from a homogeneous solution at the
irradiation wavelength. [c] pH of the solution before/after the reaction. [d] The medium pressure Hg lamp (MPL) has the most intense
emission at 365–366 nm. This type of lamp has also very intense lines at wavelengths shorter than 365–366 nm and a quasi-continuum
in the visible. [e] This commercial fluorescent lamp (coated-filtered Hg lamp) has an intense maximum emission at 350 nm, and the
intensity of the emission at 313 nm can be disregarded (2ϫ 40 watts). The irradiation time was 4 h. See Exp. Sect. for lamp details. [f]
MAP = 4-methoxyacetophenone. [g] Yield based on substrate 1 or 2. [h] MeCN/H2O = 1:1.
reaction conditions, no substitution product 3 (or 4 from
The photoreaction (unfiltered Hg lamp[13b]
) of 2
substrate 2) was observed [i.e., ET(2) = 226 kJmol–1;[13c,13d] (5 mmol) with n-C8F17I (1 mmol) and of 2 (5 mmol) with
Table 1, entries 4 and 5]. This would seem to imply that the n-C8F17Br (1 mmol) in water (30 mL) was also monitored
singlet excited state manifold is responsible for the forma- versus time in terms of product yields and pH; under these
tion of the aromatic substitution product in water, and that conditions, a steady increase of product yield was observed
the sensitized population of the triplet state of the amine as the pH of the solution decreased and the reaction
does not lead to an aromatic substitution reaction.
evolved, up to four hours of photoreaction, whereupon a
·
The n-C4F9 radical shows a clear-cut electrophilic char- plateau was reached (see the Supporting Information for a
acter in the aromatic substitution, as already reported for kinetics plot). We were unable to observe a difference be-
the addition to alkenes,[14,15] but the low regio- and chemo- tween rates of formation of 6 when 2 reacted either with n-
selectivities observed in organic solvents suggest that the C8F17I or with n-C8F17Br (see the Supporting Information).
polar effect is not the main factor in determining the high
In a typical radical addition reaction of Rf radicals de-
reactivity of perfluoroalkyl radicals toward aromatics (105– rived from CnF2n+1X (n = 4, 6, 8, 10, X = I, Br) on alkenes
106 dm3 mol–1 s–1; 2–3 orders of magnitude more reactive under various radical initiation methodologies in water, the
than alkyl radicals).[15] The enthalpic factor, which is related lower yields of addition products derived from RfBr were
to the bond energies involved, appears to be the predomi- interpreted as resulting from a higher BDE of Rf–Br com-
nant cause of the increased reactivity. The polar effect is pared to Rf–I in the radical addition mechanism.[3] The fact
considered to relate more to the polarizability than to the that aromatic substitution product yields with RfI and RfBr
polarity of a radical (the σ-perfluoroalkyl radicals are con- are comparable (see Table 1, entries 6 and 7 or 8 and 9),
sidered less polarizable and hence less sensitive to polar ef- could again suggest or support the notion that either Rf–X
fects than σ-carbon radicals).[16a]
We also subjected substrates 1 and 2 to the photoinduced event should be considered.
bond homolysis is not rate limiting, or a different initiation
(unfiltered medium pressure Hg lamp) substitution reaction
When substrates 1 or 2 (0.1 mmol) were allowed to react
with C8F17X (X = I, Br) in water, obtaining substitution with n-C4F9I (10 mmol) in water (4 mL) under 254 nm light
products 5 and 6, respectively, in yields indicated in Table 1, irradiation instead (40 W, quartz vessel, where Asubstrate
/
entries 6–9, Scheme 3.
AC4F9I = 1:10), products 3 and 4 were obtained in similar
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
3

Job/Unit: O21271
/KAP1
Date: 18-12-12 18:15:16
Pages: 12
A. B. Peñéñory, A. Postigo et al.
FULL PAPER
yields (based on substrate consumption; Table 1, entries 10 commonly activated in classical electrophilic aromatic sub-
and 11). The UV/Vis spectra of substrates 1, 2, and reagent stitution reactions.
n-C4F9I are depicted in Figure 1.
Scheme 4. Photoinduced perfluorobutylation reaction of carbazole
rings in MeCN/H2O.
Figure 1. (A) UV/Vis spectra of substrates 1 (···) and 2 (__) in H2O.
λmax [1; ε (dm3 mol–1 cm–1)] = 242 (5705 ), 287 (1145) nm; λmax [2;
When N-methylaniline (9) or N-methylnaphthylamine
(10) was subjected to photoreaction (unfiltered medium
pressure Hg lamp) with n-C4F9I in water under vigorous
stirring, the Rf-substitution products 11 and 12 were ob-
tained in lower yields (32 and 15% yields, respectively,
based on n-C4F9I; Table 1, entries 21 and 22), although the
global mass balance for the substrate was deficient (Ͻ50%).
On the other hand, in the photoreactions of primary aro-
matic amines such as aniline 13 and 1-naphthylamine 14
with n-C4F9I in water (unfiltered medium pressure Hg
lamp), no appreciable yields of substitution products were
encountered (Table 1, entries 23 and 24, respectively).
We had previously tested (unpublished results) radical
photosubstitution (254 nm) reactions in water of iodo-
benzenes with n-C4F9I, under reaction conditions in which
only the latter absorbed the light, and obtained the ipso
substitution product by a regular radical mechanism (leav-
ing group iodine radical). In halobenzenes, the initial radi-
cal adduct formed loses a halogen radical to render the sub-
stitution product. Under these latter reactions conditions,
ET to form the radical anion of the substrate or, later, the
substitution product, is not feasible.
ε (dm3 mol–1 cm–1)] = 218 (2.4ϫ105), 294 (2741) nm. (B) UV/Vis
spectrum of n-C4F9I in MeCN, λmax = 262 (2300 dm3 mol–1 cm–1)
nm.
Under these reaction conditions, and given the optical
density at 254 nm of mixtures of 1 or 2 with n-C4F9I, ap-
preciable homolysis of the C–I bond from n-C4F9–I is ex-
pected. The presence of substitution products under these
reaction conditions (i.e., 254 nm irradiation; Table 1, entries
10 and 11) might indicate that a cyclohexadienyl-type sub-
stituted radical could be involved in the substitution mecha-
nism. On the other hand, when reaction conditions were
changed (UV-absorption at 254 nm) to increasing concen-
tration of substrate 1 or 2 (2.5 mmol) to the detriment of
n-C4F9I absorption (0.5 mmol, where Asubstrate/AC4F9I
=
100:1) in water (4 mL), products 3 or 4 were nonetheless
obtained in fairly good yields (80 and 61% yields, respec-
tively; Table 1, entries 12 and 13). This would seem to indi-
cate that the reaction mechanism is initiated by a PET from
substrate 1* or 2* to RfI, and the substitution product is
formed by either an ET–PT or PT–ET sequence
(Scheme 1).
We also subjected methoxy-substituted aromatic com-
pounds (with no conventional leaving groups), such as an-
isole 15, in water, to the PET reaction with n-C4F9I, and
obtained the aromatic substitution products 16a and 16b in
85% yield. The 16a/16b isomer ratio obtained was 60:40
(Scheme 5, Table 1, entry 25).
The thermally-induced (100 °C, 4 h, stirring) reaction of
1 or 2 with n-C4F9I in deoxygenated water did not yield any
product (Table 1, entries 14 and 15, respectively). We also
attempted to induce the above reactions by exposure to ul-
trasound at 80 °C for 4 h; however, we did not observe any
product formation (Table 1, entries 16 and 17, respectively).
The photoreactions (unfiltered medium pressure Hg
lamp, Pyrex) of 1 and 2 in the presence of n-C4F9I in a
homogeneous MeCN solution, led to the formation of sub-
stitution products 3 and 4, respectively, in fairly good yields
(81 and 49%, respectively; Table 1, entries 18 and 19).
The photoreaction (unfiltered medium pressure Hg
lamp) of N-methylcarbazole 7 with n-C4F9I in MeCN/water
(1:1) (heterogeneous mixture) led to the formation of substi-
tution product 8 in fairly good yields (Scheme 4 and
Table 1, entry 20). As far as we know, the carbazole ring
has only been trifluoromethylated before through Pd-cataly-
sis with (triethylsilyl)trifluoromethane (STM) in organic
Scheme 5. Perfluorobutyl group substitution of anisole in water.
This metal-free methodology for the perfluoroalkylation
solvents.[16b] Of note is the fact that we herein report the of aromatic amines or the carbazole ring does not require
first perfluoroalkylation reaction (non-CF3) of the carb- the presence of conventional leaving groups, organocata-
azole ring. Interestingly, this substitution of the carbazole lysts, or thermal activation, and is conducted in environ-
ring takes place solely at the 3-position, which is a position mentally friendly media in heterogeneous solutions.
4
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O21271
/KAP1
Date: 18-12-12 18:15:16
Pages: 12
Aromatic Substitution of Arenes with Perfluoroalkyl Groups
Verification of the Reaction Mechanism – Intermolecular
singlet excited state energy of the aromatic amine, and the
Interactions in the Ground and Excited States and the PET last term represents the coulombic energy necessary to form
Mechanism
an ion pair with charges Z1 and Z2 in the medium of dielec-
tric constant ε at a distance r12.
We could not find a charge transfer complex (CTC) in
the UV/Vis spectra of 1 or 2 upon addition of n-C4F9I in
MeCN as solvent. A series of fluorescence quenching ex-
periments were performed with substrates 1, 2, and N-meth-
ylaniline (9) with n-C4F9I in MeCN, to obtain Stern–
Volmer quenching rate constants, ksv, in this solvent and
hence the quenching rate constants of the fluorescence (kq)
of substrates 1, 2, and 9 with n-C4F9I (for fluorescence
quenching spectra and Stern–Volmer graphs of substrates
1, 2, and substrate 9 with n-C4F9I, see the Supporting Infor-
mation).
From an overlap of the UV/Vis and fluorescence plots
of substrates 1, 2, and N-methylaniline 9, the excited state
(singlet) energies (E*) were calculated for 1*, 2*, and 9*
(see the Supporting Information). The redox potential of
the donor aromatic amines 1 and 2 were +0.71 V (vs. SCE,
in MeCN)[19] and +0.39 V, respectively.[20] The coulombic
term was taken as –0.05 eV, assuming a separation distance
of 0.8 nm.[21] The redox potential of the acceptor n-C4F9I
was taken to be –1.27 V, as an approximate measure given
in N,N-dimethylformamide (DMF).[17]
Thus, ΔGET from 2 to n-C4F9I was calculated to be
–1.78 eV (–41 kcal/mol), and that from 1 to n-C4F9I to be
–1.29 V (–29.7 kcal/mol), which are clearly highly spontane-
ous processes. For comparison, the photoinduced ET ad-
dition of 2 to furanone[20] was determined by Hoffmann
and co-workers to have a ΔGET of –0.37 eV (–8.5 kcal/mol)
in a mixture of MeCN/water; the data are summarized in
Table 2.[20] The remainder of the data for substrate 9 are
also collected in Table 2.
The quenching of fluorescence of 1 and 2 (and 9) by n-
C4F9I supports the PET mechanism. Furthermore, the
quenching rate constants (Table 2, entry 4) show a good
correlation between the oxidation peak potential of the
amine (Table 2, entry 1) and the theoretical values of their
HOMO energy calculated by using AM1 method (not
shown). We could not find evidence for exciplex formation
in the fluorescence spectra of 1 or 2 upon addition of n-
C4F9I. This suggests a concerted ET and C–I bond break-
ing step[17,18] and that the reaction proceeds from the singlet
Generally speaking, photochemical electron transfer re-
actions occur only when the charge transfer step is either
exergonic or Ͻ 5 kcal/mol.[27]
·
excited state of 2 affording its radical cation (2+·), n-C4F9 ,
and I– (A; Scheme 6 for 2).
In Table 2, measures of logkq (logarithm of rate constant
for fluorescence) quenching of amine substrates with n-
C4F9I (see the Supporting Information for Stern–Volmer
plots) of 1, 2, and 9 are given as a function of ΔGET, show-
ing that the ET processes takes place at a diffusion rate in
MeCN (in water the diffusion rate is taken as 8RT/3η,
where η is the viscosity of water taken as 8.90ϫ10–3 dynes/
cm2 or 0.890 cP at about 25 °C). A plot (not shown) of
logkq of substrates 1, 2, and 9 (data from Table 2, entry 4)
with n-C4F9I vs. ΔGET (Table 2, entry 5) demonstrates that
the ET process has a slope nearing zero, indicating a dif-
fusion-controlled process for these substrates in
Table 2. Rehm–Weller parameters for amines 1, 2, 9, and n-C4F9I.
Entry
1
2
9
n-C4F9I
1
2
3
4
5
6
7
8
ED/D+ [V]
EA/A– [V]
E* [eV]
+0.71[a]
–
+0.39
–
+0.73[b]
–
3.12
–
–1.27
3.22 (3.87)[c] 3.39
logkq [mol–1 dm–3 s–1
ΔGET [eV]
]
11.05
–1.29
–29.7
2.24[d]
0.042[d]
10.88
–1.78
–41.00
2.40[e]
0.20[f]
11.4
–1.17
–26.9
1.22[d]
0.032[d]
ΔGET [kcal/mol]
τ [ns]
φfluorescence
[a] Ref.[22], Eox (vs. SCE) = 0.83 V. [b] Ref.[23], measured in water,
vs. NHE. [c] Ref.[24], τ = 2.78 ns. [d] Ref.[25], τ measured in water
vs. NHE. [e] Ref.[26], measured in CH2Cl2. [f] Ref.[20], measured in MeCN.[28–29]
MeCN/H2O.
We performed a series of Laser Flash Photolysis (LFP)
experiments with N,N-dimethyl-1-naphthylamine 2 in the
presence of n-C4F9I in MeCN as solvent. Figure 2 depicts
the transient spectrum of the N,N-dimethyl-1-naphth-
ylamine radical cation 2·+ (B) in the presence of n-C4F9I,
and that of the triplet spectrum of 2 in the absence of n-
C4F9I (A).
The wavelength maxima in spectrum B (Figure 2, right)
at λ = 383, 440, and 675 nm, respectively, can be attributed
to the radical cation species 2·+ generated after laser pulse
excitation (355 nm) of 2 in the presence of n-C4F9I, which
is in excellent agreement with the reported spectrum of 2·+
generated after laser pulse excitation of 2 in the presence of
benzophenone (radical anion).[13a] On the other hand, in
the absence of n-C4F9I (spectrum A, Figure 2, left), the
wavelength maxima at 390 and 550 nm can be attributed to
the triplet-triplet absorption spectrum of 2 (32*), also in
Scheme 6. PET between 2 and RfX to generate radical ion pair A.
To gain further evidence for the ET mechanism (under
irradiation with a MPL) in the perfluorobutylation substi-
tution reaction of 1 and 2 in water, we employed the Rehm–
Weller relationship – see Equation (1) – to assess whether
the Gibbs energy change in the ET process (ΔGET) was
spontaneous or not.
Z1Z2
ΔGET = E(D/D+) – E(A/A–) – E* +
(1)
εr12
In Equation (1), E(D/D+) is the redox potential of the do- agreement with reported data.[13a] It is observed from com-
nor, and E(A/A–) the redox potential of the acceptor, E* the parison of spectra A and B in Figure 2, that 2·+ is a more
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5

Job/Unit: O21271
/KAP1
Date: 18-12-12 18:15:16
Pages: 12
A. B. Peñéñory, A. Postigo et al.
FULL PAPER
Figure 2. Transient UV/Vis spectra of N,N-dimethyl-1-naphthylamine 2 (0.8 mmoldm–3 in MeCN) after 0.55, 1.2, 1.8, 2.8, 4.4, and 7.6 μs
of laser pulsing at 355 nm excitation wavelength: (A) in the absence of n-C4F9I; (B) in the presence of n-C4F9I (11.6 mmoldm–3).
persistent transient (i.e., longer lived transient) than 32*. tion reaction of 2 with n-C4F9I (cf. spectrum A and B in
This makes the former a better candidate as an intermediate Figure 2).
(initiation event) in the photosubstitution reaction of
amines with n-C4F9I.
An important mechanistic aspect was also revealed by
the decrease in pH as a measure of the reaction progress
Shizuka and co-workers[13a] showed that in MeCN, trip- (Table 1, column 6, and Table 3, column 6). The initial pH
let energy transfer from triplet benzophenone (BP) to 2 pro- of the reaction mixture of 2 (5 mmol) and n-C4F9I (1 mmol)
duces triplet 32* and the corresponding triplet exciplex, was 5.5. After two-hour photoreaction (MPL, pyrex), the
whereas electron transfer took place during the deactivation pH of the solution decreased to 2.5. This would seem to
of triplet BP to yield the benzophenone anion and 2·+ cat- imply that proton loss might be involved in the reaction
ion radicals in MeCN/H2O (4:1 v/v).[13c] The authors pre- mechanism, probably through rearomatization of a carbo-
dicted schematically that added water may play an impor- cation intermediate such as those postulated in classical
tant role in altering the chemical interactions in the BP-2. aromatic electrophilic substitution reactions. The results of
Although we were unable to measure transient spectra in control pH experiments are shown in Table 3, entries 1 and
water or mixtures of water/MeCN of 2 and n-C4F9I, the 2.
claim of the participation of 2·+ cation radicals in water is
For any given solvent or solvent mixture, the cage scape
process is well-modeled as a unimolecular activated process
reasonable.[13c]
We were able to measure the quantum yield of 2·+ cation that is characterized by an exponential decay, and the cage
3
radical production using BP* in MeCN, and obtained a scape depends linearly on the solvent fluidity.[30] The time
value of φ2·+ = 0.37, which is, within error, close to the scale for cage scape is typically in the order of picoseconds.
quantum yield of fluorescence of 2, suggesting that 2·+ is Thus, the higher the fluidity, the faster the cage scape pro-
formed very efficiently under our reaction conditions.
cess.[30] One likely mechanism would involve radical substi-
Of note is the fact that these LFP transient experiments tution within the solvent cage A (Scheme 6). The in-cage
show that addition of n-C4F9I does not provoke enhance- radical cation of the substrate adds the Rf· radical within
ment of the spin-orbit coupling in 2 and that the triplet the cage, generating the cyclohexadienyl cation intermedi-
manifold is not a discrete intermediate in the photosubstitu- ate, which, upon proton loss, produces the substitution
Table 3. Experiments towards the mechanistic elucidation of the perfluoroalkylation reactions of arenes.
Solvent system [mL] Product (yield [%])[a] pHinitial/pHfinal
[b]
Entry
Synthetic method
Substrates [mmol]
[λmax irradiation (nm)]
1
2
3
4
5
6
7
8
254
254
1 (0.1), n-C4F9I (–)
2 (0.1), n-C4F9I (–)
1 (5), n-C4F9I (1), p-DNB (0.05)
2 (5), n-C4F9I (1), p-DNB (0.05)
1 (2.5), n-C4F9I (0.5)
H2O (4)
H2O (4)
H2O (30)
H2O (30)
H2O[d] (4)
H2O[d] (4)
3 (–)
4 (–)
5.5:5.5[c]
5.5:5.5[c]
5.5:2.5[c]
5.5:2.5[c]
14:14[c]
14:14[c]
14:14[c]
14:14[c]
MPL[c]
MPL[c]
254
3 (80)
4 (45)
3 (70)
4 (41)
3 (71)
4 (41)
254
254
254
2 (2.5), n-C4F9I (0.5)
1 (2.5), n-C4F9I (0.5), p-DNB (0.05) H2O[d] (4)
2 (2.5), n-C4F9I (0.5), p-DNB (0.05) H2O[d] (4)
[a] Yield based on n-C4F9I unless otherwise indicated. [b] pH of the solution before/after reaction. [c] The medium pressure Hg lamp has
the most intense emission at 365–366 nm. This type of lamp has also very intense lines at wavelengths shorter than 365–366 nm and a
quasi-continuum in the visible. [d] pH 14.
6
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O21271
/KAP1
Date: 18-12-12 18:15:16
Pages: 12
Aromatic Substitution of Arenes with Perfluoroalkyl Groups
product. We argue against in-cage product formation as be- the action of a base (H2O). Radical anion 4·– would then
ing a possible substitution route because the radical cation transfer an electron to n-C4F9I to render thermoneutral 4
of the substrate is a long-lived intermediate, as depicted in and n-C4F9· radicals, which could add to 2 to render adduct
Figure 2 (B), with sufficient time to escape the cage (1–10 μs C, continuing the propagation cycle as illustrated in
time scale). We were able to measure quantum yields for Scheme 8. This constitutes a PT–ET sequence for the rearo-
substrate 2 consumption and products 4 and 6 formation matization of the cyclohexadienyl radical intermediate (see
upon reaction of 2 with n-C4F9I and n-C8F17I, respectively, Scheme 1). A supporting piece of evidence for this mecha-
by ferrioxalate actinometry, and obtained averaged values nism would be the inhibition of the photoreaction in the
of φ–2 = 248Ϯ51 for the disappearance of substrate 2 and presence of p-dinitrobenzene (p-DNB;[31] Ered = –0.257 V
φ4 = 221Ϯ54 and φ6 = 199Ϯ65 for the formation of prod- at pH 7;[32] Ered n-C4F9I = –1.27 V), a well-known radical
ucts 4 and 6 from reaction of substrate 2 with n-C4F9I, and anion scavenger or inhibitor. After addition of p-DNB
n-C8F17I, respectively (data displayed in the table in the (0.05 mmol) to the photoreaction (unfiltered medium pres-
Supporting Information). This would clearly indicate that sure Hg lamp, Pyrex, 2 h) mixture of 1 or 2 (5 mmol) and
these substitutions proceed by a chain mechanism and that, n-C4F9I (1 mmol) in water (which represents approximately
within error, consumption of 2 leading to 4 or 6 is the only equal absorbances of substrate and p-DNB), we did not
photoprocess that takes place within the analyzed time observe any change in the yield of product 3 or 4, respec-
range.
tively (Table 3, entries 3 and 4) in comparison with experi-
We thus postulate a photoinduced ET (PET) mechanism ments performed in the absence of p-DNB (Table 1, entries
in which, upon light absorption by substrates 1 or 2, the 1 and 2, respectively). We therefore conclude that the radi-
corresponding radical cation (1+· or 2+·) together with the cal anion of 4 (i.e., 4·–) is not a discrete intermediate, as
·
n-C4F9 and iodide anion are formed in the solvent cage by illustrated in Scheme 8.
an ET (Scheme 7, for substrate 2). This being the initiation
We also conducted the photoreaction (254 nm irradia-
·
step. Upon cage-scape of 2+· (or 1+·) and n-C4F9 , the radi- tion) of n-C4F9I (0.5 mmol) and 1 or 2 (2.5 mmol) in water
·
cal n-C4F9 adds to the 4-position of 2 (and the para posi- (4 mL, 2h) at pH 14 to explore the possibility of a PT–
tion of 1) to render an aromatic substituted cyclohexadienyl ET sequence in the rearomatization of the cyclohexadienyl-
radical intermediate C (in the case of 2). Intermediate C substituted intermediate C or its analogue induced by the
donates an electron (ET) to n-C4F9I, to generate the oxid- strong basic medium. We did not obtain any appreciable
ized cation intermediate D (a σ adduct that is stabilized by change in the yield of substitution products 3 and 4
resonance effect from the N atom, i.e., a Wheland interme- (Table 3, entries 5 and 6).[31] Furthermore, these latter reac-
·
·
diate) and n-C4F9 + I–. This latter n-C4F9 triggers a chain tions were not retarded in the presence of p-DNB, indicat-
sequence. Upon proton loss (PT), adduct D generates the ing the absence of radical anion intermediates (4·– and its
substitution product 4. n-C4F9 radicals enter the substitu- analogue for substrate 1) under these extremely basic condi-
·
tion cycle depicted in Scheme 7. It should be noted that this tions (Table 3, entries 7 and 8).
·
is a chain PET mechanism in which n-C4F9 behaves as a
Probably, during direct photolysis (irradiation at 254 nm,
radical chain carrier.
conditions where Asubstrate/AC4F9I ≈ 100:1, see Table 1, entry
Another mechanistic possibility would involve the radi- 12), the initiation entails a similar PET mechanism to that
cal anion of the substitution product 4 (i.e., 4·–), generated postulated to occur at 366 nm (i.e., Scheme 7; or medium
by proton loss from intermediate C (see Scheme 8) through pressure Hg lamp), and the chain is maintained through an
Scheme 7. Proposed reaction mechanism for the photoinduced ET (PET) perfluorobutylation substitution of aromatic amines in water.
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
7

Job/Unit: O21271
/KAP1
Date: 18-12-12 18:15:16
Pages: 12
A. B. Peñéñory, A. Postigo et al.
FULL PAPER
Scheme 8. Possible reaction mechanism involving a radical anion species.
ET to n-C4F9I to generate n-C4F9· radicals. Under different n,π* singlet excited state of the amines (localization of spin
reaction conditions (irradiation at 254 nm for which distribution on N atom), these dimethyl-substituted and
Asubstrate/AC4F9I ≈ 1:10, see Table 1, entry 10) the initiation monomethyl-substituted aromatic amines, channel their re-
·
entails homolysis of n-C4F9I, to render n-C4F9 radicals activity into substitution reactions. Probably, the low mass
which enter into the propagation cycle (Schemes 7 and 8). balance obtained from N-methylaniline 9 could be attrib-
We argue in favor of a chain reaction triggered through ET uted to a deprotonation process of the radical cation 9·+, as
from radical adduct C (Scheme 7) to RfX, as evidenced by predicted by the studies of Brede and co-workers. It is likely
the lowering of pH (proton loss) in the reaction sequence that deprotonation of primary amine radical cations (pKa
(Table 1, column 6).
of aniline radical cation is 6.4[23b,23c]) might be much more
On the other hand, the radical cation 1·+ is known to accelerated in water than deprotonation of secondary amine
undergo dissociation in the presence of nucleophiles (i.e., radical cations, diverting the aromatic-ring substitution
H2OǞH3O+) to 1· radicals, which could further react in pathways in the former.[37] Then, when the proton affinity
the presence of oxygen, as shown in Equation (2), at the of the solvent becomes comparable to or larger than that
diffusion level.[33,34]
of the amine radical, this could assist in the deprotonation
of the amine radical cation.
The electron-rich anisole substrate 15 also acts as a good
electron donor to the n-C4F9I under photostimulation, as
is the case for substrates 1, 2, and 9. We surmise that, given
the oxidation potential of 15 (E(D/D+) = 1.76 V[37d]), and
its excited state (singlet) energy (E* = 86 kcal/mol)[14] [see
supplementary material for excited state (singlet) energy de-
termination], the ΔGET (15*ǞRfI) equals –0.9 eV [or –25
kcal/mol, Equation (1)], an exergonic process. Unlike the
charge transfer quenching of naphthols or phenols, which
can occur either by electron or proton transfer,[38,39] com-
pound 15 could only form a charge transfer complex by
electron transfer to RfI. However, the UV/Vis spectrum of
15, upon addition of n-C4F9I, does not present an ad-
ditional absorption maximum, and neither is an exciplex
formed.
Baciocchi and co-workers[35] have shown that α-amino
carbon radicals such as N-methylanilinomethyl radicals
shown in Equation (2), can be converted into the N-de-
methylated aniline (i.e., N-methylaniline 9) by oxidation of
N-methylanilinomethyl radicals to a carbinolamine, which,
in turn, affords the N-methylaniline and CH2O, which is
detected as its dimedone adduct.[35] Although, in their case,
the sequence of the oxidation reaction was triggered by the
use of phthalimide N-oxyl radical (PINO), which generated
the radical cation of the N,N-dialkylamine, one cannot rule
It has to be pointed out that PET from 2-naphthoxide
out a side-reaction pathway like this to account for the large ions to aliphatic halides proceeds with substitution on the
excess of substrate that we need in our substitution reac- 1-, 3-, and 6-positions of the naphthalene ring through a
tions; however, no N-methylaniline 9 was encountered in radical anion mechanism.[40a] Regarding substrate 15, pre-
the reactions that were carried out by irradiation of the sub- viously reported radical substitutions triggered by photode-
strate N,N-dimethylaniline 1 in water under our reaction composition (185 nm) of azo compounds in benzene as sol-
conditions.[36]
vent led to a different isomer distribution, i.e., ortho- (40%),
Hoffmann and co-workers[20] recently illustrated the pho- meta- (23%), and para- (37%) substitution.[40b] Minisci and
toinduced (350 nm) electron-transfer addition of N,N-di- co-workers[16a] have demonstrated that in the radical per-
methylnaphthylamines to electron-deficient alkenes in fluorobutylation reaction of various aromatic nuclei (in-
MeCN/H2O, and postulated the generation of a PET inter- cluding anisole) in tetrahydrofuran (THF) as solvent under
mediate (the radical cation of the amine) to rationalize the various radical initiation methodologies, the electrophilic
·
products obtained photochemically; this was supported by character of n-C4F9 radical is, however, barely reflected in
a deuterium isotope study.[20]
the orientation of the substitution, i.e., the ortho, meta, and
It is known that secondary amine radical cations decay para positions of aromatics are substituted with low selec-
immediately by deprotonation[34] (pKa of N-methylaniline tivity.[16a,41–43] Our methodology involving water or aque-
radical cation is 7.6[23a]). Therefore, as we postulate that the ous mixtures leads to substitution at the 4- and 2-positions
PET from 1, or 2, to n-C4F9I arises from population of the of the anisole ring, rendering a more regioselective radical
8
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O21271
/KAP1
Date: 18-12-12 18:15:16
Pages: 12
Aromatic Substitution of Arenes with Perfluoroalkyl Groups
strates were also commercially available and were previously dis-
tilled and stored over molecular sieves (4 Å) prior to use.
substitution procedure than the previously reported ap-
proaches. The result of the substitution of carbazole 7 at
the 3-position is also remarkable. It would appear that the Methods of Irradiation: An unfiltered medium pressure Hg lamp
(λmax = 365–366 nm) in water-cooled 20 or 30 mL Pyrex vessels
was used for irradiation. The temperature was maintained at 20 °C
by means of a circulating coolant liquid in the internal vessel. The
assembly was maintained 20 cm from the lamp. The water-hetero-
geneous mixtures of substrates (5 mmol) and RfX (1 mmol) were
previously deoxygenated by means of a continuous stream of Ar
for 15 min. prior to irradiation. The vessels contained a stirring
bar and the mixtures were stirred continuously by means of a mag-
netic stirrer placed underneath the vessel during the whole irradia-
tion (4 h). After photolysis, the heterogeneous mixtures were ex-
tracted into CH2Cl2 thrice and the organic layers gathered and
dried through an anhydrous Na2SO4-packed column. The extracts
were concentrated under reduced pressure, and either purified by
column chromatography through silica-gel with CH2Cl2/heptane
(50:50) or vacuum distilled. Irradiation at 254 nm (three 20 W
lamps, λ = 254 nm) was conducted in 15 mL quartz cells under
vigorous stirring. The vessels were placed 4 cm from the lamps.
Irradiation with fluorescent 350 nm Hg lamps (2ϫ 40 W were con-
ducted with two commercial Actinic BL 40 W lamps (BaSi2O5, Pb
phosphor) in a 30-mL vessel in deoxygenated water with constant
stirring during 4 h at 2 cm distance from the lamps.[43b]
orientation is related more to the stabilization of the substi-
tuted cation intermediate (electron availability of the ring)
rather than the stability of the intermediate radical adducts,
which are less sensitive to the substituent position (see
Scheme 7).
Conclusions
We have shown that aromatic amines such as N,N-di-
methylaniline, N,N-dimethyl-1-naphthylamine, N-substi-
tuted carbazoles, and methoxy-substituted aromatics react
in water by PET with RfX to render substitution products
in fairly good yields. Of note is the fact that the proposed
mechanism is a HAS mechanism in water or aqueous mix-
tures, followed by an ET and then proton transfer (PT)
steps, accounting for the observed yields of substitution
products, whereby the radical cations of the substrates are
formed in the initiation step, and a radical mechanism
seems to be superimposed on a redox process. Concerning
some mechanistic details, the reaction is compared with
radical and electrophilic aromatic substitutions.
Actinometry: The actinometry was carried out at 365 nm (unfiltered
medium pressure Hg lamp). The conversion of the actinometer (po-
tassium ferrioxalate) was monitored by measuring the absorbance
of the complex Fe–(o-phenanthroline) at 510 nm. A radiant power
of ca. 4ϫ10–9 einsteins s–1 was achieved after 20 min irradiation.
The conversion of 2 was followed by gas chromatographic tech-
niques and kept as low as 5%. Product and substrate concentra-
tions were determined relative to an internal standard (eicosane)
and were corrected for relative FID response. The global mass bal-
ance exceeded 95%. All were run in duplicate parallel experiments
and averaged over two determinations each (see Table S4 in the
Supporting Information). The sample and the actinometer (dupli-
cates) were placed equidistant from the lamp in a fixed arrange-
ment. The preparation of the actinometer, potassium ferrioxalate,
K3Fe(C2O4)3·3H2O, was performed according to the protocol sug-
gested by Hatchard and Parker.[43c]
Experimental Section
General Methods: The internal standard method was used for
quantitative GC analysis using authentic samples when available;
one of the following capillary columns was employed (1% phenyl-
SiO phase): 5 mϫ0.53 mm i.d. or 30mϫ0.32 mm i.d. Oven pro-
gram: 50 °C for 5 min then 5 °C/min to 250 °C. NMR spectra were
recorded at 400 MHz (for 1H) or 100.6 MHz (for 13C) or at
376.17 MHz (for 19F) in CDCl3 as deuterated solvent and refer-
enced to the residual solvent peak (δ = 7.26 ppm for 1H, δ =
77.0 ppm for 13C) and an external CF3COOH reference for 19F
NMR spectra, respectively. The NMR spectrometers were commer-
cial instruments: Varian Mercury 400. Hydrogen multiplicity (CH,
CH2, CH3) information was obtained from carbon DEPT-135 ex-
periments. Some NMR spectroscopic data connectivity (particu-
larly that of 10a and 10b) were confirmed by 1H-13C HSQC and
HMBC 2-D experiments. Mass spectra were acquired through di-
rect insertion (EI-technique) in a probe heated at 180 °C, monitor-
ing the ion current, by electron ionization technique (EI). FTIR
spectra were obtained with 4 cm–1 resolution of sample-impreg-
nated NaBr pellets with a commercial FTIR spectrometer Nicolet
5700 FT-IT/ATR. UV/Vis, fluorescence, and NLFP experiments
were carried out with commercial instruments Varian Cary Eclipse.
Reduced pressure distillation employed a bulb-to-bulb distillation
apparatus with four glass bulbs. When necessary, compounds were
isolated by flash chromatography performed on silica gel. High-
resolution mass spectrometry measurements were performed with
a resolution of 1–5 ppm on a ZAB-E instrument.
Time-Resolved Absorption Spectroscopy: The Laser Flash Photoly-
sis system was based on a pulsed Nd:YAG laser, using 355 nm
radiation as excitation wavelength. The single pulses were of ca.
10 ns duration, and the energy was ca. 20 mJ/pulse.
Spectroscopic Characterization of Known Compounds: These com-
pounds were purified by silica-gel column chromatography as men-
tioned above, and characterized by comparison of their 1H, 13C,
and 19F NMR spectroscopic data (when available) with those in
the literature. Addenda data below are reported for the first time
(references of compounds throughout the text).
N,N-Dimethyl-4-(1,1,2,2,3,3,4,4,4-nonafluorobutyl)benzeneamine
(3):[43] Yield 56 mg isolated (88 % GC yield) [starting from 1
1
(1 mmol), n-C4F9I (0.2 mmol)]. H NMR (300.01 MHz, CDCl3): δ
= 3.02 (s, 6 H, Me), 6.76 (d, J = 8.80 Hz, 2 H), 7.41 (d, J = 8.80 Hz,
2 H) ppm. 13C NMR (100.6 MHz, CDCl3): δ = 40.4 (Me), 111.5
(CH), 128.3 (CH), 152.8 (C) ppm. 19F NMR (376.17 MHz,
CDCl3): δ = –81.48 (CF3), –109.84 (CF2), –123.27 (CF2), –126.04
(CF2) ppm. GC/MS (EI): m/z (%) = 339 (50) [M·+], 320 (10), 170
Water was obtained from a milli-pore system, and extraction and
chromatographic solvents were HPLC-grade. Products from
Table 1, and other sections were characterized by standard spectro-
scopic techniques and compared with spectroscopic data from the
literature when available.
(100). FTIR: ν = 1618 (s), 1350 (m), 1234 (broad s), 1201 (broad
˜
s), 868 (m), 810 (m), 741 (m) cm–1.
Materials: Perfluoroalkyl iodides and bromides were commercially
available and used as received from the supplier. Aromatic sub-
N,N-Dimethyl-4-(1,1,2,2,3,3,4,4,5,5,6,6,7,7,8,8,8-heptadecafluoro-
1
octyl)benzeneamine (5):[44] Yield 463 mg (86%). H NMR (300.01,
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
9

Job/Unit: O21271
/KAP1
Date: 18-12-12 18:15:16
Pages: 12
A. B. Peñéñory, A. Postigo et al.
FULL PAPER
CDCl3): δ = 3.02 (s, 6 H, Me), 6.73 (d, J = 8.80 Hz, 2 H), 7.40 (d,
J = 8.80 Hz, 2 H) ppm. 13C NMR (100.6 MHz, CDCl3): δ = 40.1
δ = 2.98 (s, 3 H, Me), 6.77 (d, J = 9 Hz, 2 H), 7.48 (m, 2 H) ppm.
13C NMR (100.6 MHz, CDCl3): δ = 30.7, 118.4, 125.6, 149.6 ppm.
(Me), 111.3 (CH), 127.9 (CH), 152.3 (C) ppm. 19F NMR 19F NMR (376.17 MHz, CDCl3): δ = –80.32 (CF3), –110.48 (CF2),
(376.17 MHz, CDCl3): δ = –81.29 (CF3), –109.70 (CF2), –120.56 –123.45 (CF ), –125.54 (CF ) ppm. FTIR: ν = 3348 (w), 2964 (m),
˜
2
2
(CF2), –121.71 (CF2), –122.12 (CF2), –122.40 (CF2), –123.18 (CF2), 1261 (s), 1093 (br. s), 1027 (br. s), 800 (s) cm–1. GC/MS (EI): m/z
–126.61 (CF2) ppm. GC/MS (EI): m/z (%) = 539 (15) [M·+], 520
(%) = 325 (1), 310 (17), 220 (100). HRMS (EI): m/z calcd. for
C11H8F9N 325.0513; found 325.0514. C16H12F9N, calcd. C 40.63,
H 2.48, F 52.58, N 4.31; found C 40.05, H 2.71, N 4.21.
(5), 170 (100). FTIR: ν = 1618 (s), 1533 (m), 1368 (m), 1242 (br.
˜
s), 1151 (br. s), 815 (m), 655 (m) cm–1.
1-Methoxy-4-(1,1,2,2,3,3,4,4,4-nonafluorobutyl)benzene (16a):[14,44a]
Yield 11 mg isolated (85 % GC yield); yellowish oil. 1H NMR
(300.01 MHz, CDCl3): δ = 3.85 (s, 3 H, Me), 7.01 (d, J = 7.7 Hz,
2 H), 7.50 (d, J = 8.1 Hz, 2 H) ppm. 13C NMR (100.6 MHz,
CDCl3): δ = 55.3, 120.3, 120.8, 133.6, 159.7 ppm. 19F NMR
(376.17 MHz, CDCl3): δ = –81.09, –110.34, –122.38, –125.82 ppm.
Supporting Information (see footnote on the first page of this arti-
1
cle): Copies of the H, 19F, and 13C NMR spectra as well as FTIR
spectra are provided for all key intermediates and final products;
additional information as needed; LFP, UV, as well as fluorescence
spectra.
FTIR: ν = 2923 (s), 1583 (m), 1336 (m), 1203 (m), 992 (m) cm–1.
˜
Acknowledgments
1-Methoxy-2-(1,1,2,2,3,3,4,4,4-nonafluorobutyl)benzene (16b):[14,44a]
Yield 9 mg isolated (85 % GC yield); yellowish oil. 1H NMR
(300.01 MHz, CDCl3): δ = 3.85 (s, 3 H, Me), 7.01 (m, 1 H), 7.12
(m, 1 H), 7.50 (m, 2 H) ppm. 13C NMR (100.6 MHz, CDCl3): δ =
55.8, 112.4, 117.6, 128.4, 128.5, 129.2, 158.5 ppm. 19F NMR
(376.17 MHz, CDCl3): δ = –81.21, –110.75, –122.98, –125.96 ppm.
Thanks are given to financial agencies including Conicet-Argentina
and Agencia Nacional de Promoción Científica y Técnica, Ar-
gentina (PICTO-CRUP 30891) is gratefully acknowledged.
[1] a) A. Studer, D. P. Curran, Angew. Chem. Int. Ed. 2011, 50,
5018–5022; b) S. E. Vaillard, A. Studer, Radical Arylations, in:
Encyclopedia of Radicals in Chemistry, Biology and Materials
(Eds.: C. Chatgilialoglu, A. Studer), John Wiley & Sons, 2012,
vol. 2, pp. 1059–1093.
[2] A. L. J. Beckwith, V. W. Bowry, W. R. Bowman, E. Mann, J.
Parr, J. M. D. Storey, Angew. Chem. 2004, 116, 97; Angew.
Chem. Int. Ed. 2004, 43, 95–98.
[3] S. Barata-Vallejo, A. Postigo, J. Org. Chem. 2010, 75, 6141–
6148.
[4] a) M. Slodowicz, S. Barata-Vallejo, A. Vázquez, N. Nudelman,
A. Postigo, J. Fluorine Chem. 2012, 135, 137–143; b) A. Post-
igo, Can. J. Chem. 2012, 90, 493–497.
[5] C.-P. Zhang, Q.-Y. Chen, Y. Guo, J.-C. Xiao, Y.-C. Gu, Chem.
Soc. Rev. 2012, 41, 4536–4559.
[6] S. Barata-Vallejo, A. Postigo, Eur. J. Org. Chem. 2012, 1889–
1899.
[7] W. Dmowski, Z. Urbanczyk-Lipkowska, D. Wójcik, J. Fluorine
Chem. 2009, 130, 509–511.
Spectroscopic Characterization of Unknown Compounds
N,N-Dimethyl-4-(1,1,2,2,3,3,4,4,4-nonafluorobutyl)naphthalene-1-
amine (4): Yield 86 mg isolated (57% GC yield); yellowish oil. H
1
NMR (400.1 MHz, CDCl3): δ = 2.95 (s, 6 H, Me), 7.05 (d, J =
8.1 Hz, 1 H), 7.41 (t, J = 8.06, 7.97 Hz, 1 H), 7.54 (m, 3 H), 7.70
(d, J = 8.25 Hz, 1 H) ppm. 13C NMR (100.6 MHz, CDCl3): δ =
45.2 (Me), 112.4 (C), 118.3 (CH), 123.3 (CH), 124.5 (CH), 125.6
(CH), 126.1 (CH), 129.2 (C), 132.2 (C), 155.6 (C) ppm. 19F NMR
(376.17 MHz, CDCl3): δ = –81.32 (CF3), –111.64 (CF2), –123.3
(CF ), –126.17 (CF ) ppm. FTIR: ν = 2945 (m), 1581 (s), 1234 (br.
˜
2
2
s), 1098 (br. s), 806 (br. s) cm–1. GC/MS (EI): m/z (%) = 389 (95)
[M·+], 370 (18), 220 (100). HRMS (ESI): m/z calcd. for C16H12F9N:
389.0826; found 389.0876. C16H12F9N: calcd. C 49.37, H 3.11, F
43.93, N 3.60; found C 49.05, H 3.71, N 3.21.
N,N-Dimethyl-4-(1,1,2,2,3,3,4,4,5,5,6,6,7,7,8,8,8-heptadecafluoro-
octyl)naphthalene-1-amine (6): Yellowish oil, 68% GC yield (25 mg
1
isolated). H NMR (400.1 MHz, CDCl3): δH = 2.96 (s, 6 H, Me),
[8] W. Dmowski, K. Piasecka-Maciejewska, J. Fluorine Chem.
2010, 131, 746–750.
7.08 (d, J = 8.06 Hz, 2 H), 7.54 (m, 2 H), 7.71 (m, J = 8.25 Hz, 1
H), 8.18 (d, J = 7.52 Hz, 1 H) ppm. 13C NMR (100.6 MHz, [9] a) Q. Qi, Q. Shen, L. Lü, J. Fluorine Chem. 2012, 133, 115–
119; b) Q. Qi, Q. Shen, L. Lü, Chin. J. Chem. 2011, 29, 2681–
2683.
[10] H. P. Cao, J.-C. Xiao, Q.-Y. Chen, J. Fluorine Chem. 2006, 127,
1079–1082.
[11] Non-radical perfluoroalkylation of aromatic nuclei has recently
been accomplished through Pd-catalysis in benzene as solvent,
see: R. N. Loy, M. S. Sanford, Org. Lett. 2011, 13, 2548–2551.
[12] a) Y. Ohtsuka, T. Yamakawa, Tetrahedron 2011, 67, 2323–2331;
b) T. Kino, Y. Nagase, Y. Ohtsuka, K. Yamamoto, D. Uragu-
chi, K. Tokuhisa, T. Yamakawa, J. Fluorine Chem. 2010, 131,
98–105; c) S. Murakami, H. Ishii, T. Tajima, T. Fuchigami,
Tetrahedron 2006, 62, 3761–3769.
CDCl3): δC = 45.1 (Me), 115.7 (C), 124.9 (CH), 125.6 (CH), 127.4
(CH), 128.3 (CH), 128.4 (CH), 131.7 (C) ppm. 19F NMR
(376.17 MHz, CDCl3): δF = –81.16 (CF3), –107.47 (CF2), –120.53
(CF2), –121.70 (CF2), –122.12 (CF2), –122.30 (CF2), –123.09 (CF2),
–126.49 (CF ) ppm. FTIR: ν = 2962 (m), 1582 (m), 1242 (br. s),
˜
2
1210 (br. s), 802 (br. s) cm–1. GC/MS (EI): m/z (%) = 589 (7), 220
(100). HRMS (ESI): m/z calcd. for C20H12F17N 589.0698, found
589.0658. C20H12F17N calcd. C 40.76, H 2.05, F 54.81, N 2.38;
found C 40.05, H 2.61, N 2.41.
9-Methyl-1-perfluorobutyl-9H-carbazole (8): Yield 5 mg isolated
1
(50% GC yield); white solid; m.p. 72–73 °C. H NMR (500 MHz,
[13] a) M. C. Cuquerella, F. Boscá, M. A. Miranda, J. Org. Chem.
2004, 69, 7256–7261; b) the medium-pressure Hg lamp has the
most intense emission at 365–366 nm. This type of lamp has
also very intense lines at wavelengths shorter than 365–366 nm
and a quasi-continuum in the visible. We did not filter off the
less intense radiation, i.e., 313 nm, from this MPL lamp, in-
stead we used another commercial fluorescent Hg BL lamp,
with no emission at 313 nm, and maximum output at 350 nm,
as in the experiment in Table 1, entry 3, in order to test the
effect of low emission radiation in the MPL; c) T. Kiyota, M.
Yamaji, A. Shizuka, J. Phys. Chem. 1996, 100, 672–679; d) M.
Yamaji, T. Kiyota, H. Shizuka, Chem. Phys. Lett. 1994, 226,
199–205.
CDCl3): δ = 3.90 (s, 3 H, Me), 7.31 (m, J = 0.98, 7.21, 7.96 Hz, 1
H), 7.46 (d, J = 7.24 Hz, 1 H), 7.48 (d, J = 8.76 Hz, 1 H), 7.55 (m,
J = 1.14, 7.15, 8.24 Hz, 1 H), 7.66 (dd, J = 0.80, 8.7 Hz, 1 H), 8.14
(m, J = 0.80, 1.14, 7.84 Hz, 1 H), 8.31 (d, J = 1.14 Hz, 1 H) ppm.
13C NMR (125 MHz, CDCl3): δ = 26.7 (CH3), 108.9, 109.9, 119.5,
119.9, 120.0, 120.1, 120.6, 122.5, 123.9, 126.5, 141.8, 142.6 ppm.
19F NMR (470.55 MHz, CDCl3): δ = –181.71, –209.03, –223.12,
–226.22 ppm. HRMS (EI): m/z calcd. for C17H10F9N 399.0670;
found 399.0654.
N-Methyl-4-(1,1,2,2,3,3,4,4,4-nonafluorobutyl)benzeneamine (11):
1
Yield 45 mg (32%); yellowish oil. H NMR (400.1 MHz, CDCl3):
10
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O21271
/KAP1
Date: 18-12-12 18:15:16
Pages: 12
Aromatic Substitution of Arenes with Perfluoroalkyl Groups
[14] X. X. Rong, H.-Q. Pan, W. R. Dolbier Jr, J. Am. Chem. Soc.
1994, 116, 4521–4522.
Biophys. Acta Rev. Bioenerg. 1985, 811, 265–322; f) J. R. Bol-
ton, M. D. Archer, Basic Electron-Transfer Theory, in: Electron
Transfer in Inorganic, Organic, and Biological Systems (Eds.:
J. R. Bolton, N. Mataga, G. McLendon), The American Chem-
ical Society, Washington, 1991, p. 7.
[15] a) C. R. Davis, D. J. Burton, Z.-Y. Yang, J. Fluorine Chem.
1995, 70, 135–140; b) T. D. Le, R. A. Arlauskas, J. G. Weers, J.
Fluorine Chem. 1996, 78, 155–163; c) X.-T. Huang, Q.-Y. Chen,
J. Org. Chem. 2001, 66, 4651–4656; d) M. P. Scannell, D. J. Fen-
ick, S.-R. Yeh, D. E. Falvery, J. Am. Chem. Soc. 1997, 119,
1971–1977.
[16] a) A. Bravo, H.-R. Bjørsvik, F. Fontana, L. Liguori, M. Mele,
F. Minisci, J. Org. Chem. 1997, 62, 7128–7136; b) E. J. Cho,
T. D. Senecal, T. Kinsel, Y. Zhang, D. A. Watson, S. L. Buch-
wald, Science 2010, 328, 1679–1681.
[29] a) T. N. Inada, K. Kikuchi, Y. Takahashi, H. Ikeda, T. Miyashi,
J. Photochem. Photobiol. A: Chem. 2000, 137, 93–97; b) K. Ki-
kuchi, M. Hoshi, T. Niwa, Y. Takahashi, T. Miyashi, J. Phys.
Chem. 1991, 95, 38–42.
[30] A. B. Stickrath, E. C. Carroll, X. Dai, D. A. Harris, A. Rury,
B. Smith, K.-C. Tang, J. Wert, R. J. Sension, J. Phys. Chem. A
2009, 113, 8513–8522.
[17] C. P. Andrieux, L. G. Clis, M. Medebielle, P. Pinson, J. M. Sa- [31] a) R. A. Rossi, A. B. Pierini, A. B. Peñéñory, Chem. Rev. 2003,
veant, J. Am. Chem. Soc. 1990, 112, 3509–3520.
103, 71–167; b) J. I. Bardagí, V. A. Vaillard, R. A. Rossi, The
SRN1 Reaction, in: Encyclopedia of Radicals in Chemistry, Bio-
logy, and Materials (Eds.: C. Chatgilialoglu, A. Studer), Wiley,
Chichester, UK, 2012, pp. 3333–3364.
[18] We could speculate that given the heterogeneous mixture of
amines 1, or 2, and n-C4F9I in H2O, the local absorbance of the
latter (n-C4F9I) is much lower than in a MeCN homogeneous
mixture of 1, or 2, and n-C4F9I (Figure 1, B), which precludes
even more the direct homolysis of the F9C4–I bond, and re-
stricts direct homolytic substitution of the ring. The very
heterogeneity of our medium diverts the simple homolytic sub-
stitution mechanism to a radical ion chain reaction.
[32] P. Wardman, J. Phys. Chem. Ref. Data 1989, 18, 1637.
[33] J. Holcman, K. Sehested, J. Phys. Chem. 1977, 81, No. 20.
[34] A. Maroz, R. Hermann, S. Naumov, O. Brede, J. Phys. Chem.
A 2005, 109, 4690–4696.
[35] E. Baciocchi, M. Bietti, M. F. Gerini, O. Lanzalunga, J. Org.
Chem. 2005, 70, 5144–5149.
[19] A. P. Darmanyan, W. S. Jenks, J. Phys. Chem. A 1998, 102,
7420–7426.
[36] N. Vettorazzi, H. Fernandez, J. J. Silber, L. Sereno, Electro-
chim. Acta 1990, 35, 1081–1088.
[20] a) R. Jahjah, A. Gassama, F. Dumur, S. Marinkovic, S. Rich-
ert, S. Landgraf, A. Lebrun, C. Cadiou, P. Sellesz, N. [37] a) N. Karakostas, S. Naumov, O. Brede, J. Phys. Chem. A 2009,
Hoffmann, J. Org. Chem. 2011, 76, 7104–7118; b) A. G. Gries-
beck, N. Hoffmann, K. D. Warcecha, Acc. Chem. Res. 2007,
40, 128–140.
113, 14087–14094; b) C.-C. Zeng, J. Y. Becker, J. Org. Chem.
2004, 69, 1053–1069; c) S. Fukuzumi, J. Yuasa, N. Satoh, T.
Suenobu, J. Am. Chem. Soc. 2004, 126, 7585–7594; d) S. Di-
leesh, K. R. Gopidas, Chem. Phys. Lett. 2000, 330, 397–402.
[38] L. M. Tolbert, S. M. Nesselroth, J. Phys. Chem. 1991, 95,
10331–10336.
[39] a) A. R. Leheny, N. J. Turro, J. M. Drake, J. Phys. Chem. 1992,
96, 8498–8502; b) R. Pérez-Ruiz, M. A. Izquierdo, M. A. Mi-
randa, J. Org. Chem. 2003, 68, 10103–10108.
[21] A. I. Nova, C. D. Borsarelli, J. J. Cosa, C. M. Previtali, J. Pho-
tochem. Photobiol. A: Chem. 1998, 115, 43–47.
[22] A. Zweig, W. G. Hodgson, W. H. Jura, J. Am. Chem. Soc. 1964,
86, 4124–4129.
[23] a) M. Jonsson, D. D. M. Wayner, J. Lusztyk, J. Phys. Chem.
1996, 100, 17539–17543; b) A. Yu, Y. Liu, Z. Li, J.-P. Cheng,
J. Phys. Chem. A 2007, 111, 9978–9987; c) F. G. Bordwell, X.- [40] a) J. E. Argüello, A. B. Peñéñory, J. Org. Chem. 2003, 68, 2362–
M. Zhang, J.-P. Cheng, J. Org. Chem. 1993, 58, 6410–6416.
[24] I. B. Berman, Handbook of Fluorescence Spectra of Aromatic
Molecules; Academic: New York, 1971.
2368; b) T. Nakamura, Y. Koga, J. Chem. Soc. Perkin Trans. 2
1998, 659–662.
[41] H. L. Tavernier, F. Laine, M. D. Fayer, J. Phys. Chem. A 2001,
[25] S. Tobita, K. Ida, S. Shiobara, Res. Chem. Intermed. 2001, 27,
105, 8944–8957.
205–218.
[42] F. D. Lewis, R. S. Kalgutkar, Y. S. Wu, X. Y. Liu, J. Q. Liu,
R. T. Hayes, S. E. Miller, M. R. Wasielewski, J. Am. Chem. Soc.
2000, 122, 12346–12351.
[26] R. A. Lampert, L. A. Chewter, D. Phillips, Anal. Chem. 1983,
55, 68–73.
[27] L. Eberson, Electron Transfer Reactions in Organic Chemistry,
Springer-Verlag, Berlin, Germany, 1987.
[43] a) W. A. Sheppard, J. Am. Chem. Soc. 1965, 87, 2410–2420; b)
T. Justel, H. Nikol, C. Ronda, Angew. Chem. Int. Ed. 1998, 37,
3084–3103; c) C. G. Hatchard, A. Parker, J. Chem. Soc. 1956,
518–536.
[44] a) X.-T. Huang, Q.-Y. Chen, J. Org. Chem. 2001, 66, 4651–
4656; b) H. Lee, A. Czaeny, M. A. Battiste, L. Strekowski, J.
Fluorine Chem. 1998, 91, 221–224; c) T. Umemoto, Y. Kuriu,
H. Shuyama, Chem. Lett. 1981, 1663–1666.
Received: September 24, 2012
[28] a) N. Sutin, Nuclear and Electronic Factors in Electron Transfer:
Distance Dependence of Electron-Transfer Rates, in: Electron
Transfer in Inorganic, Organic, and Biological Systems (Eds.:
J. R. Bolton, N. Mataga, G. McLendon), The American Chem-
ical Society, Washington, 1991; p. 25; b) R. A. Marcus, J.
Chem. Phys. 1956, 24, 966–978; c) R. A. Marcus, J. Chem.
Phys. 1956, 24, 979–988; d) R. A. Marcus, Annu. Rev. Phys.
Chem. 1964, 15, 155–196; e) R. A. Marcus, N. Sutin, Biochim.
Published Online: ᭿
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
11

Job/Unit: O21271
/KAP1
Date: 18-12-12 18:15:16
Pages: 12
A. B. Peñéñory, A. Postigo et al.
FULL PAPER
Aromatic Fluoroalkylation
Electron-rich aromatic compounds react in
aqueous media by PET with RfX to render
substitution products in reasonable yields.
The proposed mechanism is a homolytic
aromatic substitution, followed by ET and
(PT) steps, in the absence of base; radical
cations of the substrates are formed in the
initiation step. The reaction is compared
with radical and electrophilic aromatic sub-
stitutions.
S. Barata-Vallejo, M. M. Flesia,
B. Lantaño, J. E. Argüello,
A. B. Peñéñory,* A. Postigo* .......... 1–12
Heterogeneous Photoinduced Homolytic
Aromatic Substitution of Electron-Rich
Arenes with Perfluoroalkyl Groups in
Water and Aqueous Media – A Radical-
Ion Reaction
Keywords: Heterogeneous catalysis / Green
chemistry / Radical reactions / Radical
ions / Electron transfer / Alkylation
12
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0