Regiospecific nucleophilic substitution in 2,3,4,5,6-pentafluorobiphenyl as model compound for supramolecular systems. Theoretical study of transition states and energy profiles, evidence for tetrahedral SN2 mechanism

Article abstract of DOI:10.1016/j.jfluchem.2010.09.003

2,3,4,5,6-Pentafluorobiphenyl (PFBi) was modified by the nucleophilic substitution of one fluorine using a series of O-, S- and N-nucleophiles, viz. alkaline salts of 2,2,2-trifluoro-ethanol, 3,3,4,4,5,5,6,6,7,7,8,8,8- tridecafluorooctanol, 1,2;3,4-di-O-i

Full text of DOI:10.1016/j.jfluchem.2010.09.003

Journal of Fluorine Chemistry 131 (2010) 1327–1337
Contents lists available at ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
Regiospecific nucleophilic substitution in 2,3,4,5,6-pentafluorobiphenyl as model
compound for supramolecular systems. Theoretical study of transition states and
energy profiles, evidence for tetrahedral SN2 mechanism
a,1
a,
b
´ˇ
ˇ a,1
ˇ
´
´
*
Jaroslav Kvıcala , Michal Benes , Oldrich Paleta , Vladimır Kral
a
´
Department of Organic Chemistry, Institute of Chemical Technology Prague, Technicka 5, Ustav organicke chemie Technicka, 16628 Prague 6, Czech Republic
b
´
Department of Analytical Chemistry, Institute of Chemical Technology Prague, Technicka 5, 16628 Prague 6, Czech Republic
A R T I C L E I N F O
A B S T R A C T
Article history:
2,3,4,5,6-Pentafluorobiphenyl (PFBi) was modified by the nucleophilic substitution of one fluorine using
a series of O-, S- and N-nucleophiles, viz. alkaline salts of 2,2,2-trifluoro-ethanol, 3,3,4,4,5,5,6,6,7,7,8,8,8-
tridecafluorooctanol, 1,2;3,4-di-O-isopropylidenexylitol, allylsulfane, 3,3,4,4,5,5,6,6,7,7,8,8,8-trideca-
fluorooctane-1-thiol, 3-aminopropan-1-ol (7), and tert-butyl N-(3-aminopropyl)carbamate (8). All the
substitutions took place exclusively at the position para to the phenyl group. (3-Amino-propyl)amino
derivative of PFBi (15) was further modified at the terminal amino group by acylation or fluoroalkylation.
The reaction of 8 was applied to meso-5,10,15,20-tetrakis-(pentafluorophenyl)porphyrin (20) to afford
tris- (21) and tetrakis-substituted (22) products with complete para-regioselectivity. Theoretical studies
of the reaction pathways of PFBi with ammonia, microsolvated lithium fluoride or lithium hydroxide
revealed that no Meisenheimer-type intermediates are formed in the course of the simulated reactions:
instead, tetrahedral SN2 mechanism was found. Significant regioselectivity of the nucleophilic aromatic
substitution, leading to 4-substituted products, was predicted based on relative transition state energies
in agreement with the observed experimental results.
Received 18 June 2010
Received in revised form 8 September 2010
Accepted 15 September 2010
Available online 22 September 2010
Keywords:
2,3,4,5,6-Pentafluorobiphenyl
meso-5,10,15,20-Tetrakis-
(pentafluorophenyl)-porphyrin
Nucleophilic substitution
Computational chemistry
Meisenheimer complex
Tetrahedral SN2 mechanism
ß 2010 Elsevier B.V. All rights reserved.
1. Introduction
aromatic nucleophilic substitutions of fluorine in PFBi, which
have been successfully applied in particular cases to the
Pentafluorophenyl group easily undergoes nucleophilic replace-
ment of fluorine, e.g. in 2,3,4,5,6-pentafluorobenzonitrile [1],
2,3,4,5,6-pentafluorotoluene [2,3] or 2,3,4,5,6-pentafluoro-biphenyl
[3a,4,5]. This group can be introduced to particular molecules as
macrocycles or other supermolecules and nanostructures to enable
their further modification at the peripheral parts. Typical examples
have been substituted porphyrins or metalloporphyrins bearing
meso-penta-fluorophenyl groups, in which the pentafluorophenyl
moietiesweremodifiedbyaromaticnucleophilicreplacement[6–9].
Explorational experiments with macrocyclic compounds are
usually expensive owing to high prices of the substrates.
Therefore, we were looking for a commercially accessible model
compound for meso-pentafluorophenylated porphyrin that would
reflect its chemical properties as much as possible. Such
compound has been found in 2,3,4,5,6-pentafluorobiphenyl (PFBi,
1a). We have verified on 5,10,15,20-tetrakis(pentafluorophenyl)-
porphyrin that PFBi could well simulate chemical properties of
this macrocycle [8]. In this paper we present regiospecific
pentafluorophenylated porphyrin [8].
The second advantage of PFBi as a model compound has been the
complete regioselectivity of nucleophilic displacements of fluorine
atominthissubstrate.ThenucleophilicreactionsonPFBireportedso
far, viz. reactions with O- [3a,4], S- [5a] or N-nucleophiles [5] took
place exclusively at the para-position with respect to the phenyl
group in PFBi. The same regioselectivity has been observed for
5,10,15,20-tetrakis(pentafluorophenyl)porphyrin [6–9].
2. Experimental results and discussion
An overview of the nucleophilic substitutions in PFBi studied in
this work is given in Table 1 together with the structures of the
products 9–15. All the substitutions took place with the complete
regioselectivity and exclusive nucleophilic attack at the position
para with respect to the phenyl group. No byproducts were formed
in the reactions as verified by checking the reaction mixtures by 19
NMR or following the reactions by 19F NMR.
F
2.1. Reactions of pentafluorobiphenyl with O-nucleophiles
* Corresponding author. Fax: +420 224311082/220444288.
ˇ
E-mail addresses: Jaroslav.Kvicala@vscht.cz (J. Kvı´cala), Oldrich.Paleta@vscht.cz
O-Nucleophiles used were alkoxides of the mother hydroxy
compounds, because free hydroxyls appeared to react very slowly.
(O. Paleta).
1
Fax: +420 220 444288.
0022-1139/$ – see front matter ß 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.jfluchem.2010.09.003

ˇ
J. Kvı´cala et al. / Journal of Fluorine Chemistry 131 (2010) 1327–1337
1328
Table 1
Overview of nucleophilic substitution reactions. [TD$LINE]
F
F
F
F
F
F
F
F
NuH or NuNa
or NuK
Nu
F
+
2-8
9-15
1a
.
Entry no.
Nucleophile
Product
Yield (%)
[
[TD$LINE]
F
F
F
F
1
2
9
O CH2-CF3
97
CF3-CH2-OH + K2CO3
E
[TD$LINE]
F
F
F
CF3-(CF2)4-CF2-CH2-CH2-ONa
2
3
10
38
O CH2-CH2-CF2-CF2-CF2-CF2-CF2-CF3
F
[
[TD$LINE]
O
O
F
F
F
O
O
O
3
4
11
O
CH2 CH CH CH CH2
60
O
O
O
ONa
F
F
F
[
[TD$LINE]
F
F
CH2=CH-CH2-SH + K2CO3
4
5
12
96
S CH2-CH=CH2
[
[TD$LINE]
F
F
F
F
CF3-(CF2)6-CF2-CH2-CH2-SH
+ K2CO3
5
6
13
92
S CH2-CH2-CF2-CF2-CF2-CF2-CF2-CF2-CF2-CF3
[
[TD$LINE]
F
F
F
F
+ K2CO3
H2N
OH
6
7
14
60
N
OH
H
[
[TD$LINE]
F
F
F
F
O
O
H2N
N
O
+ K2CO3
N
H
N
7
8
15
O
74
H
H

ˇ
J. Kvı´cala et al. / Journal of Fluorine Ch
[
emistry 131 (2010) 1327–1337
1329
F
F
F
In the case of 2,2,2-trifluoroethanol (2), the corresponding
alkoxylate was generated directly in the reaction mixture using
potassium carbonate in refluxing dioxane to give the desired ether
9 in 97% yield in 3 days. Nucleophilic alkoxides of 2-(perfluor-
F
F
F
F
F
F
F
H2N
dioxane, K2CO3
NHBoc
ohexyl)ethan-1-ol (3) and 1,2;3,4-di-O-isopropylidene-D-xylitol
N
H
F
F
F
(4) were prepared by their reaction with sodium hydride in
tetrahydrofuran and afforded the corresponding substitution
products 10 (38%) and 11 (60%), respectively.
N
N
H
N
F
2.2. Reactions of pentafluorobiphenyl with S- and N-nucleophiles
F
F
F
F
PFBi (1a) reacted very easily with S-nucleophiles. The
substitution reactions were carried out in N,N-dimethylformamide
in the presence of potassium carbonate at 40–50 8C. Allyl thiol (5)
afforded the sulfide 12 in a 96% isolated yield, while in the reaction
of 2-(perfluoroctyl)ethane-1-thiol (6) the resulting sulfide 13 was
obtained in a 92% isolated yield.
F
F
20
O
O
The reactions of primary amines with PFBi were performed in
the presence of potassium carbonate. 3-Aminopropanol (7) was
used in an unprotected form as the amino group appeared to be
much more reactive than the hydroxy group. The reaction in
refluxing dioxane was rather slow and afforded the expected
substitution product 14 in 60% yield in 3 weeks. The reaction
proceeded more rapidly in N,N-dimethylformamide at 50 8C to
afford the product 14 in 40% isolated yield in 5 days. Propane-1,3-
diamine was used in BOC-semiprotected form 8 and the reaction in
refluxing dioxane gave the expected product 15 in 74% yield. No
byproducts were detected in the above reactions.
HN
O
NH
O
F
F
H
HN
F
N
F
F
F
F
F
N
F
H
N
N
H
N
F
F
F
2.3. Transformations of substitution products and porphyrin
derivatives
F
F
N
R
F
F
H
The sulfide 13 was subsequently oxidized by trifluoroperox-
yacetic acid to the corresponding and more hydrophilic sulfoxide
1T 6 in 82% yield (Scheme 1). The successful preparation of the
[()D$FIG]
O
HN
21, R = F (31%)
O
H
N
F
F
O
HN
O
(30%)
CF3COOOH
H2O
22, R =
13
15
S
O
Scheme 2. Alkylaminationsinmeso-5,10,15,20-tetrakis(pentafluorophenyl)porphyrin
(20).
F
F
C8F17
F
F
16
Et2O
HCl
N
H
NH2.HCl
diamine derivative 15 offered a possibility for further transfor-
mations of the terminal amino group by fluoroacylation and
fluoroalkylation reactions (Scheme 1). First, the amino group was
deprotected in aprotic acid media [10] to give hydrochloride 17
(yield 98%). Its reaction with triethylamine in dichloromethane
liberated free amine, which was in situ acylated by perfluoro-
octanoyl chloride to afford the compound 18 in 72% isolated
yield. Fluoroalkylation of the hydrochloride 17 was carried out
F
F
17
F
F
F
F
O
N
N
CF2-CF2-CF2-CF2-CF2-CF2-CF3
H
H
18
using
2-(2,2,3,3,4,4,5,5,6,6,7,7,7-tridecafluoroheptyl)-oxirane
[11] in the presence of triethylamine. The reaction in 2,2,2-
trifluoroethanol at 60 8C gave bis-hydroxyfluoroalkylated deriv-
ative 19 in 49% yield.
NEt3
C7F15COCl
17
Our experience in the chemistry denoted in Table 1 has been
applied to meso-5,10,15,20-tetrakis(pentafluorophenyl)porphyrin
(20) as reported in our recent papers [8]. In this paper, the reaction
of 20 with Boc-semiprotected propane-1,3-diamine (8, Scheme 2)
in dioxane at 100 8C was studied: it proceeded slowly and even
after three weeks the monosubstitution of fluorine atoms in
pentafluorophenyls was not complete to afford a mixture of tris-
(21) and tetrakis-substituted (22) products in 31% and 30%
respective yields.
C6F13CH2(C2H4O)
NEt3
F
F
F
N
CF2-CF2-CF2-CF2-CF2-CF3
N
2
H
OH
F
19
Scheme 1. Transformations of the products of primary substitution in PFBi.

ˇ
J. Kvı´cala et al. / Journal of Fluorine Chemistry 131 (2010) 1327–1337
[
Nu
Nu
Nu
F
F
F
Nu
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
F
.
.
.
.
para-substitution
(-)
(-)
(-)
(-)
..
.. (-)
Nu
F
Y
Y
Y
Y
F
F
F
F
D
C
B
A
F
F
F
F
.. (-)
Y
F
F
F
F
Nu
F
F
F
F
F
F
F
..
Nu
(-)
Nu
Nu
Nu
(-)
.
(-)
.
.
(-)
.
F
meta-substitution
F
F
F
F
Y
Y
Y
Y
H
G
E
F
Y = C6H5 (1a), H (1b), F (1c), CH3 (1d), OCH3 (1e), NH2 (1f)
Scheme 3. Regioselectivity of nucleophilic substitution of fluorine in monosubstituted pentafluorobenzene moiety.
2.4. Regioselectivity of the substitutions in terms of substituent effects
and mesomeric concept
nucleophilic substitution [12], but the system is the same for all the
substituents discussed.
On the other hand, in the mesomeric (resonance) structures
F–H for the meta-substitution the substituent Y does not interact
with the alpha-negative charge. For that reason, the set F–H is
more stable (less energetic) for substituents Y with a (higher)
negative value of the sp constant than the set B–D. The result is
that in the compounds 1e or 1f (Table 2) the substituents OCH3 or
NH2 support meta-substitution. A small portion of the ortho-
substitution in 1b and 1e can be combined with a smaller steric
effect of the substituents H and OCH3 relatively to CH3, NH2 and
C6H5 [13], which block the access of a nucleophilic reagent to the
ortho-position.
PFbi (1a) can be considered as one member of the series of
substituted pentafluorobenzenes having a general structure 1 with
a substituent Y (Scheme 3). The results of the regioselectivity of
nucleophilic substitutions in the compounds 1a–1f are sum-
marised in Table 2. The data reveal that para- and/or meta- and
ortho-substitution occur in dependence on the character of the
substituent Y. Substituents attached to an aromatic system are
characterised by the Hammett sp substitutent constants. The
values of the Hammett sp constants in Table 2 exhibit that
substituents OCH3 and NH2 having negative value of sp below
ꢀ0.20 cause meta-substitution. Schematic transition states (acti-
vated complexes) for para- and meta-substitution, which may not
be Meisenheimer-type intermediates, are illustrated by the
structures A and E in Scheme 3.
3. Quantum chemical calculations of the reaction of 2,3,4,5,6-
pentafluorobiphenyl (1a) with model nucleophiles
An explanation of the para- (and ortho-) or meta-substitution
using the Hammett constants can be interpreted by the mesomeric
concept as an adequate theoretical background. Negative values of
Excluding early semiempirical studies, only a few papers can
be found dealing with computations on nucleophilic aromatic
substitution at ab initio or DFT levels. Thus, for the model
reactions of monohaloarenes with halide anions the attack of
fluoride on fluorobenzenes proceeds through Meisenheimer
complex. Moreover, this complex is a global minimum on the
PES (potential energy surface). In contrast to that, no such
complex was found for other halobenzene-halide systems [17].
Simulated reaction of pentafluoronitrobenzene with ammonia
proceeded through the corresponding intermediary Meisenhei-
mer complex as a shallow minimum for ortho substitution,
however, no such intermediate was found for the para attack
[18]. In contrast to this, simulated reactions of a series of
perfluoroarenes including perfluorobiphenyl with naked fluoride
ion gave Meisenheimer complexes which corresponded well
with experimentally observed sites of attack [19]. No transition
states were reported in this paper.
the quantities sp mean destabilisation of a negative charge. For the
para-substitution, the mesomeric (resonance) structures B–D can
be drawn (Scheme 3). In the structure B, the substituent Y interacts
with an alpha-negative charge, which is stabilised by substituents
with positive sp value. The more negative value of a sp constant,
the higher destabilisation of the mesomeric structure B. An
analogous set of mesomeric structures can be drawn for the ortho-
substitution. Fluorine substituents in the C6F5 moiety are also
active from the mesomeric point of view and support of a
Table 2
Regioselectivity of the SN reactions according the Scheme 3 and Hammett sp
constants.
Y
Regioselectivity of SN
Ref.
Hammetta sp
3.1. Calculations with ammonia as a model nitrogen nucleophile
const.
1a
1b
C6H5
H
para
[3a,4,5]
[14]
0.02
0.00
Strong preference for the substitution of PFBi (1a) with
nucleophiles at the position 4 was observed experimentally. To
understand better the reasons for the high regioselectivity, we
decided to study computationally aromatic nucleophilic substitu-
tion on PFBi. As a first model nucleophile, we chose ammonia as the
simplest nitrogen nucleophile and as the respective centres of
nucleophilic attack C-3 and C-4 carbons of the PFBi (1a) skeleton.
Our computational study started with geometry optimization of
para (95–97%) + ortho
(3–5%)
1c
1d
1e
F
SN
[15b,c]
[2,3]
0.06
CH3
OCH3
para
ꢀ0.14
ꢀ0.28
para (52%) +ortho
(16%) +meta (36%)
meta (79–90%) + para
(10–21%)
[15a]
1f
NH2
[12,16]
ꢀ0.57
a
Ref. [13].

ˇ
G
J. Kvı´cala et al. / Journal of Fluorine Ch
I
emistry 131 (2010) 1327–1337
1331
F
F
F
F
B
C
H
H
5
3
4
H3N
F
H
A
H
H
1a
Fig. 1. Considered nucleophilic attacks of ammonia on PFBi (1a): due to the C2
symmetry of the substrate, front face and rear face attacks of ammonia on C(4) are
equivalent (path A), while front face attack on C(5) is equivalent with rear face
attack on C(3) (path B) and front face attack on C(3) is equivalent with rear face
attack on C(5) path (C).
PFBi. The calculated torsion angle between both aromatic rings is
468 as a result of compromise between attempts to preserve at
least some aromaticity and steric hindrance of the ortho-
substituents in good agreement with a 538 angle found in the
crystal structure [20]. Following that, three nucleophilic attacks
were considered in the calculations, i.e. pathways A (at C-4), B (at
less hindered meta side denoted here C-5) and C (at more hindered
meta side denoted C-3) (Fig. 1).
The reaction pathway study started with finding equilibrium
geometries. We employed DFT method at the PBE1PBE/6-31+G(d)
level of theory (Fig. 2, Table 3). We generally prefer PBE1PBE
hybrid functional over the common B3LYP functional, as the
former contains only one non-empirical parameter (hence this
functional is also named PBE0) and has better description of
systems with non-covalent interactions [21]. Due to the formation
of fluoride anion in the course of the reaction, we also included a
single set of diffuse functions into the basis set used to achieve
correct description of anionic intermediates.
Search for saddle points revealed in analogy to Ref. [18] no
intermediary Meisenheimer complexes on the PES, the corre-
sponding Meisenheimer structures being transition states as
confirmed by vibrational analysis. IRC calculations of both sides
of all three reaction pathways A–C (Fig. 2) starting from the
corresponding transition states did not find any intermediary local
minima between the depicted transition states and educts/
products clusters (Fig. 2).
Fig. 2. Energies and structures of saddle points of the reaction of pentafluorobiphenyl
(1a) with ammonia (PBE1PBE/6-31+G(d)).
We did not find any minimum with ammonia complexed to the
pentafluorobiphenyl molecule (1a) at the C-3 for pathway C due to
preferred coordination of ammonia to ortho-hydrogen of the
pentafluorobiphenyl system, which leads to a local minimum
lower by ca. 4 kJ molꢀ1 than other precoordinated educts.
Inspection of all three reaction pathways A–C proved that the
reaction regioselectivity is driven kinetically as the barrier of the
reverse reaction exceeds 200 kJ molꢀ1, too high for making the
reaction reversible under the experimental conditions compared
to ca. 130 kJ molꢀ1 for the forward reaction. In agreement with the
energies of published Meisenheimer intermediates [19], we found
strong preference for the attack of ammonia at the C-4 position of
Table 3
Relative energies (kJ molꢀ1) of educts, products and transition states of reaction of pentafluorobiphenyl (1a) with ammonia by various methods, role of inclusion of zero point
energy (ZPE) and thermal correction (GFE, 298.15 K) into the calculations.
Method
Energy
Path C
Geometry
P
Path A
Geometry
E
Path B
Geometry
E
TS
E
TS
P
TS
P
I
Erel.
2.3
230.9
15.1
81.7
85.6
215.8
0.0
0.0
86.0
228.3
12.5
1.3
ETSrel.
Erel.
ETSrel.
Erel.
ETSrel.
Erel.
ETSrel.
Erel.
ETSrel.
Erel.
ETSrel.
Erel.
ETSrel.
Erel.
ETSrel.
Erel.
II
2.4
3.5
2.6
2.8
2.3
2.5
1.6
0.4
250.3
15.8
82.8
60.4
76.9
99.2
76.9
70.3
95.9
91.9
85.7
60.8
78.4
100.9
79.2
64.0
94.3
92.1
234.5
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
86.3
61.3
79.1
101.3
79.5
69.5
94.7
97.0
247.8
13.2
1.4
2.7
1.5
2.4
1.4
1.8
3.6
5.5
III
IV
V
310.0
15.0
292.7
0.0
307.7
17.3
255.9
15.9
240.0
0.0
253.5
13.6
230.3
21.2
209.1
0.0
228.3
19.2
I
232.2
12.8
219.4
0.0
229.7
10.3
E + ZPE
I
240.2
13.0
227.1
0.0
237.8
10.7
GFE
V
227.2
19.3
207.9
0.0
224.9
17.0
E + ZPE
V
234.4
19.7
214.7
0.0
231.8
17.1
GFE
ETSrel.
A: C-4 substitution; B: C-5 substitution (less hindered); C: C-3 substitution (more hindered); P: products; TS: transition state; E: educts; I: PBE1PBE/6-31+G(d)//PBE1PBE/6-
31+G(d); II: PBE1PBE/6-311++ G(2df,p)//PBE1PBE/6-31+G(d); III: MP2/6-311++G(2df,p)//PBE1PBE/6-31+G(d); IV: BMK/6-311++G(2df,p)//PBE1PBE/6-31+G(d); IEF-
PCM(THF)BMK/6-311++G(2df,p)//IEFPCM(THF)-BMK/6-311++G (2df,p); E: electronic energy; E + ZPE: electronic energy + zero point energy; GFE: Gibbs free energy.
V

ˇ
J. Kvı´cala et al. / Journal of Fluorine Chemistry 131 (2010) 1327–1337
1332
PFBi, the difference in the transition state energy being ca.
12 kJ molꢀ1, which corresponds to more than 99% excess of the
para-substituted product (Table 3).
indeed obtained using our level of theory (PBE1PBE/6-31+G(d))
comparable results with global minimum being the corresponding
Meisenheimer complex of the C-4 substitution and local minimum
its analogue for the C-3 substitution. However, apart of the
Meisenheimer complexes, no other minima on the potential energy
surface (PES) of the system, as well as no transition states
corresponding to the formation of these complexes could be found
(see Supporting info). Similar results, i.e. preference for the C-4
attack over other positions and non-existence of other saddle points
other than that between the Meisenheimer complexes were
obtained for simulated reaction of naked fluoride ion with PFBi (1a).
To our opinion, these results indicate that the use of naked
fluoride anion for calculations of simulated aromatic nucleophilic
substitution leads to completely incorrect description of PES of the
studied system. Especially, naked fluoride anion displays too high
reactivity and hence neither the fluoride ion-fluoroarene cluster,
not the transition state between this structure and Meisenheimer
complex could be detected. In agreement with this, fluoride ion in
ethereal solvents exist mostly as a tight ion pair with the
corresponding cation. We hence decided to add lithium cation
to the fluoride anion to complete the modified nucleophile.
Moreover, due to characteristic tetracoordination of lithium cation,
we included microsolvation to it in the form of two dimethyl ether
molecules. The respective fourth coordination site of the lithium
cation was implied to be occupied by the substrate.
In contrast to the computations including naked fluoride ion,
search for saddle points of the reaction of PFBi in analogy to the
ammonia–PFBi system using modified nucleophile–2DME-LiF,
proceeded smoothly. First, both precomplexed educt clusters and
the complexed product clusters were found quite easily, second, no
Meisenheimer complexes were detected on the PES, the correspond-
ingstructuresbeingthetransitionstates(Fig. 3,Table5).Thereaction
pathways were again checked by IRC calculations for the presence of
local minima (Fig. 3, Table 4). As the nucleophile and leaving group
are identical in this case, the simulated reaction pathway of the C-4
attack is symmetrical. Moreover, the forward reaction of micro-
solvated fluoride ion at C-3 is identical with the reverse reaction at C-
5 and, correspondingly, only two reaction pathways were studied.
Similarly tothe ammoniaas the model nucleophile, wewerenot able
tofindtheminimumcorrespondingtotheeduct-substrateclusterfor
theC-3attackduetopreferablecoordinationofthenucleophiletothe
3.2. Role of computational method, ZPE and thermochemical
correction
To evaluate
a quality of the correlation description we
recalculated the energies of all saddle points in the A–C (Fig. 2)
pathways using three various approaches. First, we employed the
same PBE1PBE functional with larger triple zeta basis set and more
polarization functions (PBE1PBE/6-311++G(2df,p)), and then
compared the results from DFT method with those obtained by
simple perturbation treatment (MP2/6-311++G(2df,p)). Further
we recalculated the energies using the BMK functional (BMK/6-
311++G(2df,p)), which is especially tailored for the transition state
calculations and gives results comparable or better than much
more demanding multiconfigurational methods [22]. Finally, as
the reactions were performed mostly in THF, we reoptimized all
saddle points using the BMK/6-311++G(2df,p) method in
a
simulated solvent using an IEF-PCM approach [23]. Although the
actual values of the activation energies change with the method
employed, the relative energies of the transition states for the
pathways A–C vary only by a few kJ molꢀ1. Thus, using a larger
basis set in the PBE1PBE method enhances a little the absolute
values of the transition state energies, MP2 method probably
overestimates the transition state energies and BMK method
affords values higher than other DFT methods but lower than MP2.
Inclusion of the solvent effects reduces the absolute values of the
transition state energies to the levels of the simplest method used
(PBE1PBE/6-31+G(d)), but results in even higher preference for C-4
substitution. In all cases, the relative energy of the transition state
of the A pathway is at least 10 kJ molꢀ1 smaller than the remaining
two transition states energies implying more than 98% excess of
the C-4 substituted product. The best method employed rises the
transition state energy to more than 19 kJ molꢀ1 (Table 4, Fig. 1 in
Supporting info). Finally, we compared for both optimization
method used, (PBE1PBE/6-31+G(d) and IEFPCM(THF)-BMK/6-
311++G (2df,p)), pure electronic energies, electronic energies
including zero point energy (ZPE) and Gibbs free energies (GFE).
The results show that considering the ZPE lowered the correspond-
ing relative energy barriers of pathways B and C (Fig. 2) by about
2 kJ molꢀ1 for both methods used, while the correction to room
temperature and other thermochemical parameters (vibration,
rotation, translation, entropy) did not change these values
significantly (see Table 4).
[)(D$FIG]
oT rtho-hydrogen of the phenyl group.
3.3. Calculations with microsolvated fluoride ion
For the model reactions with fluoride ion, to obtain results
comparable withthe most recentpublished data [19] we startedthis
part of computational study by the search for saddle points of the
reaction of decafluorobiphenyl with naked fluoride anion. We
Table 4
Relative energies (kJ molꢀ1) of educts, products and transition states of reaction of
pentafluorobiphenyl (1a) with 2DME-LiF cluster at the PBE1PBE/6-31+G(d) level of
theory.
Energy
Path A
Geometry
E
Path B
Geometry
E
TS
P
TS
P
Erel.
8.5
93.7
8.5
8.5
107.1
0.0
ETSrel.
0.0
13.4
A: C-4 substitution; B: C-5 substitution (less hindered). P: products; TS: transition
state; E: educts.
Fig. 3. Energies and structures of saddle points of reaction of pentafluorobiphenyl
(1a) with 2DME-LiF cluster (PBE1PBE/6-31+G(d)).

ˇ
J. Kvı´cala et al. / Journal of Fluorine Chemistry 131 (2010) 1327–1337
1333
Table 5
Relative energies (kJ molꢀ1) of educts, products and transition states of reaction of pentafluorobiphenyl (1a) with 2DME-LiOH cluster at the PBE1PBE/6-31+G(d) level of
theory.
Energy
Path C
Geometry
P
Path A
Geometry
E
Path B
Geometry
E
TS
E
TS
P
TS
P
Erel.
17.3
189.5
144.5
153.6
177.3
14.2
154.1
189.4
0.0
ETSrel.
12.2
0.0
12.0
A: C-4 substitution; B: C-5 substitution (less hindered); C: C-3 substitution (more hindered); P: products; TS: transition state; E: educts.
In analogy to the case of ammonia, substitution at C-4 proved to
be energetically more favored with the corresponding transition
state energy lower by more than 13 kJ molꢀ1 whereas no such
preference could be observed for the educts cluster.
phenyl moiety. High activation energy of the reverse reaction
(more than 160 kJ molꢀ1) again indicates that the reaction is driven
kinetically.
All equilibrium geometries (Cartesian coordinates) and their
energies (hartree) are listed in Supporting information.
3.4. Calculations with microsolvated hydroxide ion
3.5. Attempted explanation of regioselectivity
Following a successful description of the attack of the PFBi
with fluoride anion, we used analogous approach for the
reaction with hydroxide ion. Again, naked hydroxide anion is
unrealistic in ethereal solvents used and its use in calculations
led to the respective products at the C-3, C-4 and C-5 positions
as the only saddle point which could be detected on the PES.
However, addition of lithium microsolvated with two molecules
of DME to the hydroxide anion led to the simulated nucleophile
which allowed to identify educts clusters, transition states and
products clusters as the respective saddle points on the PES
(Fig. 4, Table 5).
Microsolvated hydroxide ion preserves high reactivity and
hence the reactions proceed through early transition state with
low activation energy. Transition state of the C-4 attack (pathway
A) is again most favored by more than 12 kJ molꢀ1 in close analogy
to both previously studied systems (NH3, 2DME.LiF). Similarly to
the previous systems, neither educts cluster of the C-3 attack nor
products cluster of the C-5 attack could be found due to preferred
Looking for the explanation of the high regioselectivity of
nucleophilic aromatic substitution in PFBi, we turned our
attention to two possible information sources, viz. analysis of
charges on the individual reaction centres in the respective educt
clusters and transition states, and analysis of frontier orbitals of
the transition states.
Analysis of the atomic charges is not a simple issue in
quantum chemistry and several approaches can be employed. As
standard Mulliken atomic charges are generally regarded
unreliable, we employed Merz–Singh–Kollman scheme based
on ESP analysis [24]. In agreement with analysis of PFBi, no
significant differences in atomic charges on the carbon centres
in pentafluorophenyl part were found for the ammonia–PFBi
educt clusters corresponding to pathways A–C (Figs. 2 and 5).
HT owever, this situation changed remarkably for the respective
[()D$FIG]
Educts:
[()D$FIG]
cT oordination of the nucleophile to the ortho-hydrogen of the
-0.74 F
F
-0.62 F
F
-0.58 F
F
0.03
F
H3N
0.15
0.09
F
0.14
0.08 F
0.04
0.14 F
0.06
F
F
0.13
NH3
F
F
F
H3N
C
B
A
Transition states:
-0.95 F
F
-0.56 F
F
0.19
-0.60F
F
0.55
0.61
0.26
0.34
0.25
F
NH3
F
H3N
F
F
0.54 F
F
F
F
H3N F
A
B
C
Fig. 5. Selected charges calculated according to the Merz–Singh–Kollman scheme
for the ammonia–PFBi educt clusters and transition states corresponding to
pathways A–C.
Fig. 4. Energies and structures of saddle points of reaction of pentafluorobiphenyl
(1a) with 2DME-LiOH cluster (PBE1PBE/6-31+G(d)).

ˇ
[
J. Kvı´cala et al. / Journal of Fluorine Chemistry 131 (2010) 1327–1337
4. Conclusions
Aromatic nucleophilic substitution reactions of PFBi (1a) with a
series of (long-chain) fluorinated O-, S- and N-nucleophiles were
studied as model reactions for supramolecular systems. The
reactions proceeded with the complete para-regioselectivity with
respect to the phenyl group to afford the products 9–15.
Some of the primary substitution products were subsequently
transformed, viz. fluoroalkyl sulfide 13 was oxidized to the
corresponding sulfoxide 16, while (3-aminopropyl)amino deriva-
tive 15 was bis-fluoroalkylated at the primary amino group to
afford dendrimeric 19. The experience in modification reactions of
1a was successfully applied to meso-5,10,15,20-tetrakis(penta-
fluorophenyl)porphyrin (20) as reported recently [8] and in
reactions with semi-protected propane-1,3-diamine (products
20 and 21).
Fig. 6. Electrostatic potential mapped on isoelectronic surface for the ammonia–
DFT calculations of the attack of model nucleophiles, ammonia,
fluoride and hydroxide anion were successfully accomplished and
based on the respective transition state energies confirmed the
experimentally observed regioselectivity. Compared to the previ-
ously used naked fluoride anion, microsolvated LiF and LiOH as the
model nucleophiles led to more realistic description of the model
chemistry.
Based on the computational analysis, the regioselectivity can be
attributed to higher polarizability of the system during the C-4
attack and improved stability of the corresponding dearomatized
transition states, which corresponds well with their lower HOMO
orbital energy. No Meisenheimer-type intermediates were formed
in the course of the simulated reactions; instead, a tetrahedral SN2
mechanism was found.
PFBi transition states corresponding to pathways A and C.
transition states. Thus, for the pathway A the charge on the
attacked carbon C-4 reaches 0.54, while for the pathways B and
C the values are 0.25 for C-5 and 0.34 for C-3. In agreement with
this the charges on C-1 (phenyl substituted, which bears highest
negative charge in educts) reach ꢀ0.95, ꢀ0.56 and ꢀ0.60 for the
respective pathways A–C (Fig. 5). This indicates that for the
most advantageous C-4 attack the
p-electrons on the penta-
fluorophenyl ring are most polarizable facilitating the nucleo-
philic attack. The C-4 carbon of the PFBi (1a) seems to be the
most prone to the dearomatization (the level of dearomatization
can be estimated from the positive charge of the attacked
carbon) and we assume that this is the major factor governing
the regioselectivity.
Visualization of the electrostatic potential for transition states
of pathways A and C (Fig. 6) revealed that some additional
stabilization of the pathway A can be attributed by some electron
flow to the partially conjugated phenyl ring.
5. Experimental
5.1. General
Analysis of frontier orbitals revealed that the HOMO orbital of
the C-4 attack transition state has significantly lower energy (by
0.007 hartree, i.e. 18 kJ molꢀ1) than that of the C-3 attack, which
corresponds well with the respective transition state energies. The
HOMO of the transition state of the more favorable attack also
spreads over the second aromatic ring in striking contrast to the
Temperature data are uncorrected. Commercially available
starting materials, reagents and solvents were used without
further purification unless otherwise stated. THF was dried by
sodium benzophenone ketyl and distilled prior to use. Column
chromatography was performed on silica gel (63–200
mm, Merck).
All reactions were carried out under nitrogen and reaction
mixtures were stirred (magnetic spinbar). Apparatus were dried
in oven and cooled under nitrogen. NMR spectra were recorded on
a Bruker 400 AM (19F at 376.6 Hz) and Varian Gemini 300 HC (1H at
300.07 Hz, 13C at 75.46 Hz) instruments: TMS and CFCl3 as the
[()D$FIG]
oT ther HOMO orbital (Fig. 7).
internal standards, chemical shifts
d in ppm, coupling constants J
(Hz), solvents CDCl3 or CD3OD. Mass spectra were scanned on a
Hewlett Packard MSD 5971A (EI 70 eV) and MALDI-TOF Biflex IV
(Bruker Daltonics) instruments.
5.2. Computations
DFT calculations were performed using Gaussian 03W
program suite [25]. MP2 and IRC calculations were accomplished
with Firefly QC package [26], which is partially based on the
GAMESS (US) source code [27]. Vibrational frequencies were
calculated for all species to characterise them as minima or
transition states. Transition states geometries were found
starting from the corresponding minima on the potential energy
surfaces and estimated transition states using Schlegel’s QST2 or
QST3 method [28]. In all cases, the connections of the
corresponding transition states and minima were verified by
IRC calculations of the simulated reaction pathways. Vizualiza-
tions of the molecules were performed with the GaussView
program [29].
Fig. 7. HOMO orbitals of the ammonia-fluoroarene 1a transition states.

ˇ
J. Kvı´cala et al. / Journal of Fluorine Chemistry 131 (2010) 1327–1337
1335
5.3. The chemicals prepared and obtained
4.14, 4.3, 4.4, 4.41 (5 ꢁ m, 7 H, xylitol), 7.44 (m, 5 H, Ar) ppm. 19F
3
5
NMR (CDCl3):
d
= ꢀ145.3 (dd, JF,F = 22.3 Hz, JF,F = 8.8 Hz, 2 F,),
3
5
3,3,4,4,5,5,6,6,7,7,8,8,8-Tridecafluorooctan-1-ol (Atofina – Elf
ꢀ157.5 (dd, JF,F = 22.3 Hz, JF,F = 8.8 Hz, 2 F) ppm. C23H24F4O5
Atochem); 1,2;3,4-di-O-isopropylidene-D-xylitol (4) was prepared
(456,4): calcd. C 60.6, H 5.34; found C 60.6, H 5.22.
according to Refs. [30a,b]; allyl thiol (5) was prepared from allyl
mercaptane and thiourea according to Ref. [30c] in 85% yield.
3,3,4,4,5,5,6,6,7,7,8,8,9,9,10,10,10-Heptadecafluorodecane-1-thiol
(6) was prepared from 3,3,4,4,5,5,6,6,7,7,8,8,9,9,10,10,10-heptade-
cafluoro-1-iododecane and thiourea according to Ref. [30d] in 44%
yield. tert-Butyl-N-(3-aminopropyl)carbamate (7) was prepared
from di-tert-butyl dicarbonate and propane-1,3-diamine in 98%
yield using a modified procedure [30e]. 2-(3,3,4,4,5,5,6,6,7,7,8,8,8-
Tridecafluoroheptyl)-oxirane was prepared from tridecafluorohexyl
iodide (Atofina – Elf Atochem) according to Ref. [30f].
5.7. 4-Allylsulfanyl-2,3,5,6-tetrafluorobiphenyl (12)
A mixture of PFBi (1a, 150 mg, 0.61 mmol), allylthiol (5, 45 mg,
0.61 mmol), K2CO3 (0.85 g, 0.61 mmol) and DMF (5 mL) was stirred
at 40 8C for 2 h. The solvent was removed under reduced pressure
(rotary evaporator) and crude product was purified by column
chromatography (hexane) and crystallised (hexane). Product 12
(172 mg, 96%) as white crystals, m.p. 72–73 8C. 1H NMR (CDCl 3):
3
3
d
= 3.58 (d, JH,H = 7.2 Hz, 2 H, CH2–S–Ar), 5.05 (dd, JH,H = 9.6 Hz,
2JH,H = 1.3 Hz,
1 H, CH = CH2, Hcis), 5.06 (dd, JH,H = 16.8 Hz,
3
3
5.4. 2,3,5,6-Tetrafluoro-4-(2,2,2-trifluoroethoxy)biphenyl (9)
2JH,H = 1.3 Hz, 1 H, CH = CH2, Htrans), 5.83 (ddt, JH,H = 9.6 Hz,
3JH,H = 16.8 Hz, 3JH,H = 7.2 Hz, 1 H, CH = CH2), 7.46 (m, 5 H, Ar) ppm.
3
5
A mixture of PFBi (1a, 50 mg, 0.205 mmol), 2,2,2-trifluoroetha-
nol (2, 2 mL, 2.74 g, 27.4 mmol), K2CO3 (42 mg, 0.305 mmol) and
dioxane (5 mL) was refluxed for 3 days. The solvent was removed
under reduced pressure (rotary evaporator) and the crude product
was purified by column chromatography (hexane) and crystallised
(hexane). Product 9 (66 mg, 97%) as white crystals, m.p. 103–
19F NMR (CDCl3):
d
= ꢀ134.4 (dd, JF,F = 24 Hz, JF,F = 12 Hz, 2 F),
ꢀ144.3 (dd, 3JF,F = 24 Hz, 5JF,F = 12 Hz, 2 F) ppm. C15H10F4S (298.3):
calcd. C 60.4, H 3.38; found C 60.4, H, 3.54.
5.8. 2,3,5,6-Tetrafluoro-4-[(3,3,4,4,5,5,6,6,7,7,8,8, 9,9,10,10,10-
heptadecafluorodec-1-yl)sulfanyl]biphenyl (13)
104 8C. 1H NMR (CDCl3):
d
= 4.57 (q, 3JHF = 8 Hz, 2 H, CH2,), 7.45 (m,
3
5 H, Ar) ppm. 19F NMR (CDCl3):
d
= ꢀ75.4 (t, JH,F = 8 Hz, 3 F),
A mixture of PFBi (1a, 50 mg, 0.205 mmol), thiol 6 (99 mg,
0.205 mmol), K2CO3 (28 mg, 0.205 mmol) and DMF (5 mL) was
stirred at 50 8C for 2 h. The solvent was removed under reduced
pressure (rotary evaporator) and crude product was purified by
column chromatography (hexane) and crystallised (hexane).
Product 13 (133 mg, 92%) as white crystals, m.p. 108–109 8C. 1H
3
5
ꢀ144.4 (dd, JF,F = 17.6 Hz, JF,F = 9.1 Hz,
2
F), ꢀ157.4 (dd,
5
3JF,F = 17.7 Hz, JF,F = 8.9 Hz, 2 F) ppm. 13C NMR (CDCl3):
d
= 70.5
2
1
(q, JC,F = 36 Hz, CH2), 122.7 (q, JC,F = 277 Hz, CF3), 128.7, 129.2,
130.1 (3x s, Ar–H), 126.8 (s, Ar–H), 145.8, 142.6, 139.5, 136.3 (4x m,
Ar–F) ppm. MS (EI, Mr = 324), m/z = 325 (10) [M++1], 255 (71), 233
(19), 135 (100).
NMR (CDCl3):
d = 2.46 (m, 2 H, CH2–RF), 3.18 (m, 2 H, CH2–S–Ar),
7.44 (m, 5 H, Ar) ppm. 19F NMR (CDCl3):
d
= ꢀ80.9 (t, 3JF,F = 10 Hz, 3
5.5. 2,3,5,6-Tetrafluoro-4-(3,3,4,4,5,5,6,6,7,7,8,8,8-tridecafluoroctan-
F, CF3), ꢀ113.9 (m, 2 F, CF2–CH2), ꢀ121.7 (m, 6 F, 2 ꢁ CF2), ꢀ122.5
1-yloxy)biphenyl (10)
(qs, 2 F, CF2), ꢀ123.0 (qs, 2 F, CF2), ꢀ125.9 (qs, 2 F, CF2), ꢀ133.7 (dd,
3JF,F = 23.2 Hz, JF,F = 11 Hz,
2
F), ꢀ142.1 (dd, JF,F = 24 Hz,
5
3
A solution of 3,3,4,4,5,5,6,6,7,7,8,8,8-tridecafluorooctan-1-ol
(196 mg, 0.45 mmol) in THF (2 mL) was added dropwise to a
mixture of NaH (60%, 17 mg, 0.41 mmol) and THF (2 mL) while
stirring. The reaction mixture was then stirred at r.t for 0.5 h and at
40 8C 0.5 h. A solution of PFBi (1a, 100 mg, 0.41 mmol) in THF
(2 mL) and added dropwise to the solution of the sodium alkoxide
3 and the mixture was stirred at 80 8C for 2 days. The solvent was
removed under reduced pressure (rotary evaporator) and the
crude product was purified by column chromatography (hexane)
and crystallised (hexane). Product 10 (102 mg, 38%) as white
5JF,F = 11 Hz, 2 F) ppm. 13C NMR (CDCl3):
d
= 25.7 (s, SCH2), 32.5
2
(t, JC,F = 22.3 Hz, CH2RF), 107.6–122.0 (m, 7 ꢁ CF2, CF3), 127.0 (s,
Ar–H), 128.7, 129.48, 130.1 (3 ꢁ s, Ar–H), 142.7, 145.3, 146.1, 148.5
(4 ꢁ m, Ar–F) ppm. C22H9F21S (704.4): calcd. C 37.5, H 1.29; found
C 37.3, H 1.46.
5.9. 3-[N-(2,3,5,6-Tetrafluorobiphenyl-4-yl)amino]propan-1-ol (14)
A mixture of PFBi (1a, 62 mg, 0.254 mmol), 3-aminopropan-1-
ol (7, 19 mg, 0.254 mmol), K2CO3 (35 mg, 0.254 mmol) and dioxane
(5 mL) was refluxed for 2 weeks. The solvent was removed under
reduced pressure (rotary evaporator) and crude product was
purified by column chromatography (CH2Cl2) and crystallised
(hexane). Product 14 (73 mg, 60%) as white crystals, m.p. 68–70 8C.
crystals, m.p. 84–86 8C. 1H NMR (CDCl 3):
d = 2.70 (m, 2 H, CH2–RF),
3.75 (m, 2 H, CH2–O–Ar), 7.44 (m, 5 H, Ar) ppm. 19F NMR (CDCl3):
3
d
= ꢀ81.3 (t, JF,F = 10 Hz, 3 F, CF3), ꢀ113.8 (m, 2 F, CF2–CH2),
ꢀ122.4 (m, 2 F, CF2), ꢀ123.3 (m, 2 F, CF2), ꢀ124.0 (m, 2 F, CF2),
ꢀ126.6 (m, 2 F, CF2), ꢀ144.9 (dd, 3JF,F = 22.2 Hz, 5JF,F = 8.8 Hz, 2 F),
1H NMR (CDCl3): = 1.90 (qi, 3JH,H = 6 Hz, 2 H, CH2CH2CH2), 3.59 (tt,
d
3
5
ꢀ157.7 (dd, JF,F = 22.3 Hz, JF,F = 8.8 Hz, 2 F) ppm. C20H9F17
O
3JH,H = 6.6 Hz, 5JF,H = 1.65 Hz, 2 H, CH2NH), 3.84 (t, 3JH,H = 6.1 Hz, 2
(588.3): calcd. C 40.8, H 1.54; found C 40.9, H 1.68.
H, CH2OH), 7.45 (m, 5 H, Ar) ppm. 19F NMR (CDCl3):
d
= ꢀ147.2 (d,
3
3JF,F = 14 Hz, 2 F), ꢀ157.4 (d, JF,F = 14.4 Hz, 2 F,) ppm. 13C NMR
4
5.6. 1,2;3,4-Di-O-4-isopropylidene-5-O-(2,3,5,6-tetrafluorobiphenyl-
4-yl)- -xylitol (11)
(CDCl3):
d
= 32.8 (s, CH2CH2CH2), 43.8 (t, JC,F = 4 Hz, ArNHCH2),
3
D
61.1 (s, CH2OH), 127.2 (t, JC,F = 12 Hz, Ar–H), 128.2, 128.4, 130.6
(3 ꢁ s, Ar–H), 136.1, 139.4, 142.80, 146.0 (4 ꢁ m, Ar–F), 156.4 (s,
C55O) ppm. MS (EI, Mr = 299): m/z = 298 (100) [M+ꢀ1], 281 (25),
255 (88), 241 (38), 227 (38), 207 (13).
Diisopropylidenexylitol 4 (105 mg, 0.45 mmol) in THF (2 mL)
was added dropwise to a mixture of NaH (60%, 17 mg, 0.41 mmol)
and THF (2 mL). The reaction mixture was refluxed for 2.5 h. A
solution of PFBi (1a, 100 mg, 0.41 mmol) in THF (2 mL) was added
dropwise to the reaction mixture, which was then refluxed at r.t.
for 2 days. The solvent was removed under reduced pressure
(rotary evaporator) and the crude product was purified by column
chromatography (hexane–acetone, 7:1) and crystallised (hexane).
The yield of 11 was 111 mg (60%) as white crystals, m.p. 94–96 8C.
5.10. tert-Butyl N-{3-[N-(2,3,5,6-tetrafluorobiphenyl-4-
yl)amino]propyl} carbamate (15)
A mixture of PFBi (1a, 332 mg, 1.36 mmol), carbamate 8
(317 mg, 1.84 mmol), K2CO3 (188 mg, 1.36 mmol) and dioxane
(5 mL) was refluxed for 2 weeks. The solvent was removed under
reduced pressure (rotary evaporator) and crude product was twice
1H NMR (CDCl3):
d
= 1.41, 1.46, 1.49 (3 ꢁ s, 12 H, 4x CH3), 3.95,

ˇ
J. Kvı´cala et al. / Journal of Fluorine Chemistry 131 (2010) 1327–1337
1336
purified by column chromatography (hexane and hexane – CH2Cl2,
1:1) and crystallised (hexane). Product 15 (402 mg, 74%) as white
the organic layer dried over MgSO4. After removing the solvent, the
residue was chromatographed (petroleum ether–dichloro-
methane, 1:1). Product 18 (53 mg, 72%) as white crystals, m.p.
crystals, m.p. 126–127 8C. 1H NMR (CDCl3):
d
= 1.46 (s, 9 H, CH3),
1.77 (qi, 3JH,H = 6.6 Hz, 2 H, CH2CH2CH2), 3.27 (q, 3JH,H = 6.6 Hz, 2 H,
119–121 8C. 1H NMR (CDCl3):
d
= 1.92 (qi, JH,H = 6.6 Hz, 2 H,
3
3
3
OCONHCH2), 3.47 (q, JH,H = 6 Hz, 2 H, ArNHCH2CH2), 4.42, 4.64
CH2CH2CH2), 3.58, 3.51 (2 ꢁ q, JH,H = 6.6 Hz, 4 H, NHCH2), 4.11
(2 ꢁ broad s, 2 H, 2 ꢁ NH), 7.45 (m, 5 H, Ar) ppm. 19F NMR (CDCl3):
(broad s, ArNH), 6.64 (broad s, NHCO), 7.4 (m, 5 H, Ar) ppm. 19F
3
d
= ꢀ147.2 (d, 3JF,F = 14.8 Hz, 2 F), ꢀ160.8 (d, 3JF,F = 14 Hz, 2 F) ppm.
NMR (CDCl3):
d
= ꢀ81.2 (t, JF,F = 10 Hz, 3 F, CF3) ꢀ119.8 (t,
13C NMR (CDCl3):
d
= 28.4 (s, CH3), 31.3 (s, CH2CH2CH2), 37.4 (s,
3JF,F = 12.8 Hz, 2 F, CF2CO), ꢀ122.0 (m, 2 F, CF2), ꢀ122.4 (m, 2 F, CF2),
ꢀ122.9 (m, 2 F, CF2), ꢀ123.2 (m, 2 F, CF2), ꢀ126.6 (m, 2 F, CF2),
ArNHCH2), 42.8 (s, OCONHCH2), 79.5 (s, O-C), 127.0 (s, Ar–H),
128.2, 128.4, 130.3 (3 ꢁ s, Ar–H), 136.3, 139.5, 142.8, 146.2 (4 ꢁ m,
Ar–F), 156.4 (s, C = O) ppm. MS (EI, Mr = 398): m/z = 397 (35)
[M+ꢀ1], 323 (42), 265 (12), 233 (33), 156 (72), 133 (100).
3
3
ꢀ146.9 (d, JF,F = 13.7 Hz, 2 F, Ar), ꢀ160.8 (d, JF,F = 14.9 Hz, 2 F,
Ar) ppm. 13C NMR (CDCl3):
ArNHCH2), 43.0 (s, CONHCH2), 106.9–118.3 (m, 6x CF2, CF3), 126.5
d = 29.7 (s, CH2CH2CH2), 37.6 (s,
3
2
C
20H22F4N2O2 (398.4): calcd. C 60.3, H 5.57, N 6.93; found C
(t, JC,F = 11.5 Hz, Ar–H), 158.1 (t, JF,F = 26 Hz, C55O), 128.0 (s, Ar–
H), 128.3, 128.5, 130.2 (3 ꢁ s, Ar–H), 136.9, 138.9, 143.5, 145.4
(4 ꢁ m, Ar–F) ppm. C23H13F19N2O (694.3): calcd. C 39.8, H 1.89, N
4.03; found C 39.4, H 2.12, N 3.91.
60.3, H 5.64, N 7.04.
5.11. 2,3,5,6-Tetrafluorobiphenyl-4-yl
(3,3,4,4,5,5,6,6,7,7,8,8,9,9,10,10,10-heptadecafluorodec-1-yl)
sulfoxide (16)
5.14. N-{3-[N-(2,3,5,6-Tetrafluorobiphenyl-4-yl)amino]propyl}-1,10-
nitrilobis-(4,4,5,5,6,6,7,7,8,8,9,9,9,9-tridecafluorononan-2-ol) (19)
Dry flask was charged under nitrogen with CH2Cl2 (2 mL),
trifluoroacetic acid (5 mL) and sulfide 13 (50 mg, 0.071 mmol). The
solution was cooled to 0 8C, diluted H2O2 was then dropwise added
(30%, 0.1 mL) and the mixture stirred for 1 h. The mixture was
evaporated to dryness and crude product purified by column
chromatography (hexane) to get pure 16 (42 mg, 82%) as white
A mixture of amine hydrochloride 17 (20 mg, 0.085 mmol),
fluoroalkyloxirane (Scheme 1, 300 mg, 0.798 mmol), triethylamine
(9 mg, 0.089 mmol) and 2,2,2-trifluoroethanol (1 mL) was heated at
60 8C for 2 days while stirring. Volatile components were then
removed (rotary evaporator, 45 8C, 50 mmHg) and the residue was
chromatographed (petroleum ether–dichloromethane, 4:1) Product
crystals, m.p. 121–122 8C. 1H NMR (CDCl 3):
RF), 3.4 (m, 1 H, CH2–S(O)–Ar), 3.78 (m, 1H, CH2–S(O)–Ar), 7.5 (m, 5
d = 2.74 (m, 2 H, CH2–
19 (41 mg, 51%) as waxy material. 1H NMR (CDCl3):
d = 1.83 (m, 2 H,
3
H, Ar) ppm. 19F NMR (CDCl3):
d
= ꢀ81.2 (t, JF,F = 10 Hz, 3 F, CF3),
CH2CH2CH2), 2.12–2.40 (m, 4 H, CH2–RF), 2.59–2.80 (m, 6 H, CH2N),
3.50 (m, 4 H, NHCH2, 2 ꢁ CH–OH), 4.21 (m, 3 H, ArNH, 2 ꢁ OH), 7.40
ꢀ113.7 (m, 2 F, CF2–CH2), ꢀ122.1 (m, 2 F, CF2), ꢀ122.4 (m, 4 F,
2 ꢁ CF2), ꢀ123.2 (m, 2 F, CF2), ꢀ123.5 (m, 2 F, CF2), ꢀ126.6 (m, 2 F,
(m, 5H, Ar) ppm. 19F NMR (CD3Cl):
d
= ꢀ81.4 (t, 3JF,F = 9.4 Hz, CF3),
CF2), ꢀ140.5 (m, 4 F, Ar) ppm. 13C NMR (CDCl3):
d
= 45.3 (s,
ꢀ112.9 (qs, 2 F, CF2–CH2), ꢀ122.3 (qs, 2 F, CF2), ꢀ123.4 (qs, 2 F, CF2),
2
S(O)CH2), 25.5 (t, JC,F = 22.4 Hz, CH2RF), 107.6–122.0 (m, 7 ꢁ CF2,
CF3), 126.1 (s, Ar–H), 128.9, 129.9, 130.1 (3 ꢁ s, Ar–H), 142.7,
145.3, 145.3, 146.1, 148.5 (4 ꢁ m, Ar–F) ppm. MS (MALDI,
Mr = 720,35): m/z = 721.19 [M++1].
ꢀ124.1 (qs, 2 F, CF2), ꢀ126.7 (qs, 2 F, CF2), ꢀ147.0 (dd, 3JF,F =14.3 Hz,
3
5JF,F = 6.4 Hz, 2 F), ꢀ161.0 (d, JF,F = 18.2 Hz, 2 F) ppm. 13C NMR
(CDCl3):
d = 29.7 (s, CH2CH2CH2), 37.0 (s, ArNHCH2), 44.0 (m, CH2–
RF), 53.4 (s, CH2CH2N), 60.5 (s, NCH2CH(OH)), 63.9 (s, CH–OH),
107.6–121.5 (m, 12 C, 10 ꢁ CF2, 2 ꢁ CF3), 127.2 (m, Ar–H), 128.3,
128.42, 130.3 (3 ꢁ s, Ar–H), 136.6, 139.0, 143.3, 145.7 (4 ꢁ m, Ar–
F) ppm. MS (MALDI, Mr = 1050,52): m/z = 1051.43 [M++1].
5.12. 3-[N-(2,3,5,6-Tetrafluorobiphenyl-4-
yl)amino]propylammonium chloride (17)
Dry gaseous HCl was introduced into the solution of carbamate
15 (100 mg, 0.251 mmol) of Et2O (15 mL) for 1 h. The mixture was
evaporated to dryness (rotary evaporator, 40 8C, 20 mmHg) and
then dried on oil pump for 5 h to afford product 17 (83 mg, 98%) as
5.15. 5,10,15-Tris(4-{[3-(tert-butoxycarbonyl)amino]propylamino}-
2,3,5,6-tetrafluorophenyl)-20-(2,3,4,5,6-pentafluorophenyl)-
21H,23H-porphyrin (21) and 5,10,15,20-tetrakis(4-{[3-(tert-
butoxycarbonyl)-amino]propylamino}-2,3,5,6-tetrafluorophenyl)-
21H, 23H-porphyrin (22).
white crystals, m.p. 205–207 8C. 1H NMR (CD3OD):
d
= 1.97 (qi,
H,
3
3JH,H = 7.1 Hz,
2
H, CH2CH2CH2), 3.04 (t, JH,H = 7.7 Hz,
2
3
CH2NH2ꢂHCl), 3.51 (t, JH,H = 6.6 Hz, 2 H, ArNHCH2CH2), 7.4 (m, 5
A
mixture of 5,10,15,20-tetrakis(pentafluorophenyl)por-
H, Ar) ppm. 19F NMR (CD3OD):
d
= ꢀ147.5 (d, JF,F = 14 Hz, 2 F),
phyrin (20, 250 mg, 0.257 mmol), (3-amino-propyl)carbamate 8
(194 mg, 1.129 mmol), K2CO3 (156 mg, 1.129 mmol) and dioxane
(6 mL) was heated at 100 8C for 3 weeks while stirring. After that
time, TLC (hexane – acetone – CH2Cl2, 6:2:1) indicated only 2
products. The solvent was removed (rotary evaporator, 45 8C,
100 mmHg) and the residue chromatographed (solvent as for TLC)
to afford tris-substituted product 21 (113 mg, 31%), m.p. 98–
100 8C, and tetrakis-substituted product 22 (122 mg, 30%), m.p.
104–107 8C.
3
ꢀ161.0 (d, JF,F = 14 Hz, 2 F) ppm. 13C NMR (CD3OD):
d = 30.2 (s,
3
CH2CH2CH2), 38.7 (s, ArNHCH2), 43.5 (s, CH2NH2ꢂHCl), 127.02 (s,
Ar–H), 129.4, 129.5, 131.4 (3 ꢁ s, Ar–H), 137.4, 140.6, 144.2, 147.3
(4 ꢁ m, Ar–F) ppm. MS (EI, Mr = 335): m/z = 299 (75) [M++ 1–HCl],
282 (35), 254 (100), 234 (8).
5.13. N-{3-[N-(2,3,5,6-tetrafluorobiphenyl-4-yl)amino]propyl}-
2,2,3,3,4,4,5,5,6,6,7,7,8,8,8-pentadecafluoroctanamide (18)
21: 1H NMR (CDCl3): = 9.03 (d, 3JH,H = 4.5 Hz, 2 H, (-H), 9,01 (s,
d
Dry flask (25 mL) was charged with amine hydrochloride 17
(25 mg, 0.107 mmol) and CH2Cl2 (10 mL) through septum. The
solution was cooled to 0 8C and a solution of perfluorooctanoyl
chloride (92 mg, 0.212 mmol) in CH2Cl2 (3 mL) and then a solution
of triethylamine (32 mg, 0.317 mmol) in CH2Cl2 (3 mL) were added
dropwise while stirring. The mixture was stirred at 0 8C for 2 h to
become homogeneous and then evaporated to dryness (rotary
evaporator, 40 8C, 650 mmHg). The solid residue was dissolved in
Et2O (15 mL), washed with saturated water solution of KHCO3 (3ꢁ
25 mL), then with saturated water solution of NaCl (3ꢁ 25 mL) and
4 H, (-H), 8.87 (d, 3JH,H = 4.5 Hz, 2 H,
b
-H), 4.79 (t, 3JH,H = 6 Hz, 6 H,
3
6 ꢁ NH), 3.74 (t, JH,H = 6.6 Hz, 6 H, ArNHCH2CH2), 3.44 (qa,
3JH,H = 5.5 Hz, 6 H, OCONHCH2), 1.97 (qi, JH,H = 6.6 Hz, 6 H,
3
CH2CH2CH2), 1.53 (s, 27 H, CH3), ꢀ2.83 (s, 2 H, NH) ppm. 19F
3
NMR (CDCl3):
d
= ꢀ137.0 (d, JF,F = 21.5 Hz, 6 F), ꢀ161.3 (d,
3JF,F = 19.7 Hz, 6 F), ꢀ141.1 (m, 2 F, ortho), ꢀ152,8 (m, 1 F, para),
ꢀ162,5 (m, 2 F, meta) ppm. 13C NMR (CDCl3, characteristic signals):
d
= 160.1 (s, C55O), 79.65 (s, C–O), 42.9 (s, OCONHCH2), 36.7 (s,
ArNHCH2), 31.6 (s, CH2CH2CH2), 28.4 (s, CH3) ppm. MS (FAB,
Mr = 1437.28): m/z = 1437.2 [M+].

ˇ
J. Kvı´cala et al. / Journal of Fluorine Chemistry 131 (2010) 1327–1337
1337
22: 1H NMR (CDCl3):
d
= 8.96 (s, 8 H, (-H), 4.83, 4.76 (2 ꢁ broad
(c) J.P.C. Tome´, M.G.P.M.S. Neves, A.C. Tome´, J.A.S. Cavaleiro, A.F. Mendonca, I.N.
Pegado, R. Duare, M.L. Valdeira, Bioorg. Med. Chem. 13 (2005) 3878–3888;
(d) D. Samaroo, C.E. Soll, L.J. Todaro, C.M. Drain, Org. Lett. 8 (2006) 4985–4988;
(e) P.S.S. Lacerda, A.M.G. Silva, A.C. Tome´, M.G.P.M.S. Neves, A.M.S. Silva, J.A.S.
Cavaleiro, A.L. Llamas-Saiz, Angew. Chem. Int. Ed. 45 (2006) 5487–5491.
[10] (a) B.H. Lee, M.J. Miller, J. Org. Chem. 48 (1983) 24–31;
(b) A. Malabarba, R. Ciabatti, J. Kettenring, R. Scotti, G. Candiani, J. Med. Chem. 35
(1992) 4054–4060.
3
s, 8 ꢁ NH), 3.75 (qa, JH,H = 6.6 Hz, 8 H, ArNHCH2CH2), 3.44 (qa,
3JH,H= 6.0 Hz, 8 H, OCONHCH2), 1.99 (qi, JH,H = 6.6 Hz, 8 H,
3
CH2CH2CH2), 1.52 (s, 36 H, CH3), ꢀ2.84 (s, 2 H, NH) ppm. 19F NMR
(CDCl3):
d
= ꢀ141.0 (d, 3JF,F = 17.8 Hz,
8
d
F), ꢀ161.3 (d,
3JF,F = 16.1 Hz, 8 F) ppm. 13C NMR (CDCl3):
= 156.5 (s, C = O),
[11] B. Guyot, B. Ameduri, B. Boutevin, J. Fluorine Chem. 74 (1995) 233–240.
[12] (a) J. Burdon, Tetrahedron 21 (1965) 3373–3380;
147.8, 145.9, 137.9, 136.1 (m, ArF), 129.1 (m), 104.8 (m), 79.5 (s, C-
O), 42,9 (s, OCONHCH2), 37.5 (s, ArNHCH2), 31.5 (s, CH2CH2CH2),
28.4 (s, 12 C, CH3) ppm. MS (FAB, Mr = 1591.5): m/z = 1591.6 [M+].
(b) O. Paleta, Usp. Khim. 40 (1971) 855–917;
(c) R.D. Chambers, Fluorine in Organic Chemistry, A. Rowe, Chippenham, 1973,
p. 283;
(d) G.M. Brooke, J. Fluorine Chem. 86 (1997) 1–76;
(e) R.D. Chambers, P.A. Martin, G. Sandford, D.L.H. Williams, J. Fluorine Chem. 129
(2008) 998–1002.
Acknowledgments
[13] (a) J. Shorter, in: N.B. Chapman, J. Shorter (Eds.), Advances in Linear Free Energy
Relationships, Plenum Press, London, 1972, pp. 71–96;
The authors thank the Atofina (Atochem) Company for the gift
of perfluoroalkyl iodides. The research was supported by the Grant
Agency of the Czech Republic (project No. 203/01/1311) and the
Ministry of Education of the Czech Republic (project No. MSM
6046137301).
(b) O. Exner, in: N.B. Chapman, J. Shorter (Eds.), Correlation Analysis in Chemis-
try—Recent Advances, Plenum Press, New York, 1978, pp. 439–467;
(c) O. Exner, Correlative Relationships in Organic Chemistry, SNTL, Prague,
(1981), pp. 70–91.
[14] (a) P. Robson, M. Stacey, R. Stephens, J.C. Tatlow, J. Chem. Soc. (1960) 4754–4760;
(b) J. Burdon, P.L. Coe, C.R. Marsh, J.C. Tatlow, Tetrahedron 22 (1966) 1183–1188;
(c) M.E. Peach, A.M. Smith, J. Fluorine Chem. 4 (1974) 399–408.
[15] (a) J. Burdon, W.B. Hollyhead, J.C. Tatlow, J. Chem. Soc. (1965) 5152–5158;
(b) J.A. Godsell, M. Stacey, J.C. Tatlow, Nature 178 (1956) 199–200;
(c) W.J. Pummer, L.A. Wall, Nature 127 (1958) 643–644.
[16] (a) J.G. Allen, J. Burdon, J.C. Tatlow, J. Chem. Soc. (1965) 6329–6337;
(b) G.G. Yakobson, V.D. Shteingarts, G.G. Furin, N.N. Vorozhtsov, Zh. Obshch.
Khim. 34 (1964) 3514–3519.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.jfluchem.2010.09.003.
References
[17] M.N. Glukhovtsev, R.D. Bach, S. Laiter, J. Org. Chem. 62 (1997) 4036–4046.
[18] K. Tanaka, M. Deguchi, S. Iwata, J. Chem. Res. (S) (1999) 528–529.
[19] M. Muir, J. Baker, J. Fluorine Chem. 126 (2005) 727–738.
[20] C.P. Brock, D.G. Naae, N. Goodhand, T.A. Hamor, Acta Crystrallogr. B34 (1978)
3691–3696.
[1] (a) G. Chandra, A.D. Jenkins, M.F. Lappert, R.C. Srivastava, J. Chem. Soc. A
(1970) 2550–2558;
(b) L.S. Kobrina, G.G. Furin, G.G. Yakobson, Zh. Org. Khim. 6 (1970) 512–520;
(c) L.N. Markovskii, G.G. Furin, Yu.G. Shermolovich, G.G. Yakobson, Izv. Akad. Nauk
SSSR, Ser. Khim. (1981) 867–869;
(d) S. Fujii, Y. Maki, H. Kimoto, J. Fluorine Chem. 43 (1989) 131–144;
(e) V.M. Vlasov, I.A. Oskina, Russ. J. Org. Chem. 30 (1994) 1587–1592;
(f) J.H. Marriott, A.M.M. Barber, R.M. Grimshaw, S. Neidle, M. Jarman, J. Chem.
Soc., Perkin Trans. 1 24 (2000) 4265–4278.
[21] C. Adamo, V. Barone, J. Chem. Phys. 110 (1999) 6158–6170.
[22] (a) A.D. Boese, J.M.L. Martin, J. Chem. Phys. 121 (2004) 3405–3416;
ˇ
(b) J. Hajduch, O. Paleta, J. Kvı´cala, G. Haufe, Eur. J. Org. Chem. (2007) 5101–5111.
[23] (a) M.T. Cance`s, B. Mennucci, J. Tomasi, J. Chem. Phys. 107 (1997) 3032–3041;
(b) B. Mennucci, J. Tomasi, J. Chem. Phys. 106 (1997) 5151–5158;
(c) B. Mennucci, E. Cance`s, J. Tomasi, J. Phys. Chem. B 101 (1997) 10506–10517;
[2] (a) C. Tamborski, E.J. Soloski, J. Org. Chem. 31 (1966) 746–749;
(b) R. Filler, N.R. Ayangar, W. Gustowski, H.H. Kang, J. Org. Chem. 34 (1969)
534–538;
`
(d) J. Tomasi, B. Mennucci, E. Cances, J. Mol. Struct. (Theochem) 464 (1999)
211–226.
[24] (a) B.H. Besler, K.M. Merz Jr., P.A. Kollman, J. Comp. Chem. 11 (1990) 431–439;
(b) U.C. Singh, P.A. Kollman, J. Comp. Chem. 5 (1984) 129–145.
[25] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman,
J.A. Montgomery Jr., T. Vreven, K.N. Kudin, J.C. Burant, J.M. Millam, S.S. Iyengar, J.
Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.A. Petersson, H.
Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T.
Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J.E. Knox, H.P. Hratchian, J.B.
Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev,
A.J. Austin, R. Cammi, C. Pomelli, J.W. Ochterski, P.Y. Ayala, K. Morokuma, G.A.
Voth, P. Salvador, J.J. Dannenberg, V.G. Zakrzewski, S. Dapprich, A.D. Daniels, M.C.
Strain, O. Farkas, D.K. Malick, A.D. Rabuck, K. Raghavachari, J.B. Foresman, J.V.
Ortiz, Q. Cui, A.G. Baboul, S. Clifford, J. Cioslowski, B.B. Stefanov, G. Liu, A.
Liashenko, P. Piskorz, I. Komaromi, R.L. Martin, D.J. Fox, T. Keith, M.A. Al-Laham,
C.Y. Peng, A. Nanayakkara, M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen,
M.W. Wong, C. Gonzalez, J.A. Pople, Gaussian 03, Revision D.01, Gaussian, Inc.,
Wallingford, CT, 2004.
[26] A.A. Granovsky, Firefly version 7.1.G, www.http://classic.chem.msu.su/gran/fire-
fly/index.html.
[27] M.W. Schmidt, K.K. Baldridge, J.A. Boatz, S.T. Elbert, M.S. Gordon, J.H. Jensen, S.
Koseki, N. Matsunaga, K.A. Nguyen, S. Su, T.L. Windus, M. Dupuis, J.A. Montgom-
ery, J. Comput. Chem. 14 (1993) 1347–1363.
[28] P.Y. Ayala, H.B. Schlegel, J. Chem. Phys. 107 (1997) 375–384.
[29] R. Dennington II, T. Keith, J. Millam, K. Eppinnett, W.L. Hovell, R. Gilliland,
GaussView, Version 3.09, Semichem, Inc., Shawnee Mission, KS, 2003.
[30] (a) N. Baggett, K.W. Buck, A.B. Foster, R. Jefferis, B.H. Rees, J.M. Webber, J. Chem.
Soc. (1965) 3382–3387;
(c) V.M. Vlasov, O.V. Zakharova, Zh. Org. Khim. 11 (1975) 785–793.
[3] (a) R. Bolton, J.P.B. Sandall, J. Chem. Soc., Perkin Trans. 2 (1978) 746–750;
(b) R. Filler, R.C. Rickert, J. Fluorine Chem. 18 (1981) 483–496;
(c) V.M. Vlasov, G.G. Yakobson, Zh. Org. Khim. 17 (1981) 2192–2201;
(d) P.L. Coe, D. Oldfield, J.C. Tatlow, J. Fluorine Chem. 29 (1985) 341–348.
[4] (a) P.J.N. Brown, M.T. Chaudry, R. Stephens, J. Chem. Soc. C (1969) 2747–2750;
(b) D.M. Allen, A.S. Batsanov, G.M. Brooke, S.J. Lockett, J. Fluorine Chem. 108
(2001) 57–68.
[5] (a) M.T. Chaudry, R. Stephens, J. Chem. Soc. (1963) 4281–4283;
(b) D.G. Holland, J. Org. Chem. 29 (1964) 1562–1565;
(c) T.W. Lin, D.G. Naae, Tetrahedron Lett. (1978) 1653–1656.
[6] (a) P. Battioni, O. Brigaud, H. Desvaux, D. Mansuy, T.G. Traylor, Tetrahedron Lett.
32 (1991) 2893–2896;
(b) N. Jagerovic, A. Fruchier, J. Elguero, J. Heterocyclic Chem. 32 (1995) 1829–
1831;
(c) P. Battioni, E. Cardin, M. Louloudi, B. Scho¨lhorn, G.A. Spyroulias, D. Mansuy,
T.G. Taylor, Chem. Commun. (1996) 2037–2038;
(d) D. Bouy-Debec, O. Brigaud, P. Leduc, P. Battioni, D. Mansuy, Gazz. Chim. Ital.
126 (1996) 233–237;
(e) A. Sen, K. Avijit, Tetrahedron Lett. 37 (1996) 5421–5424.
[7] (a) M. Suarez, E. Salfran, R.I. Curiel, F. Gaudemer, J. Elguero, Bull. Soc. Chim. Belg.
106 (1997) 323–326;
(b) R. Breslow, B. Gabriele, J. Yang, Tetrahedron Lett. 39 (1998) 2887–2890;
(c) S.J. Shaw, C. Edwards, R.W. Boyle, Tetrahedron Lett. 40 (1999) 7585–7586;
(d) S. Evans, J.R.L. Smith, J. Chem. Soc., Perkin Trans. 2 (2000) 1541–1551.
[8] (a) B. Dolensky´, V. Kra´l, in: Proceedings of the 24th International Symposium on
Macrocyclic Chemistry, Barcelona PS2-15, 1999;
(b) L. Zarif, J. Greiner, S. Pace, J.G. Riess, J. Med. Chem. 33 (1990) 1262–1269;
(c) B. Ambekar, Y. Sarvottam, Synth. Commun. 29 (1999) 3477–3486;
(d) J.D. Glennon, S. Hutchinson, S.J. Harris, A. Walker, M.A. McKervey, C.C.
McSweeney, Anal. Chem. 69 (1997) 2207–2212;
(e) D. Muller, I. Zeltser, G. Bitan, C. Gilon, J. Org. Chem. 62 (1997) 411–416;
(f) V. Cı´rkva, B. Ame´duri, B. Boutevin, O. Paleta, J. Fluorine Chem. 83 (1997)
151–158.
ˇ
´
(b) M. Benes, O. Paleta, V. Kral, Chem. Listy 95 (2001) 743;
ˇ
(c) K. Lang, V. Kra´l, P. Kapusta, P. Kuba´t, P. Vasek, Tetrahedron Lett. 43 (2002)
4919–4922.
[9] (a) P. Pasetto, X. Chen, C.M. Drain, R.W. Franck, Chem. Commun. (2001) 81–82;
(b) R. Weiss, F. Pu¨ hlhofer, N. Jux, K. Merz, Angew. Chem. Int. Ed. 41 (2002) 3815–
3817;