New nucleophilic rearrangement in the mechanism of the three-component domino cyclisation affording fluoroalkylated (pyrrolo)quinazolines

Article abstract of DOI:10.1016/j.jfluchem.2013.10.017

The following mechanism steps were verified for the three-component domino cyclisation affording (pyrrolo)quinazolines from 2-(aminomethyl)aniline, a very reactive oxo compound and "usual" oxo compound. The first step was a rapid reaction of very reactive oxo compound (trifluoropyruvate or hexafluoroacetone) with benzylic amino group to form hemiaminal, but not imine; the second step was the reaction of oxo compound with aromatic amino group to form imine (Schiff base) being in equilibrium with its enamine form; the third step was an intramolecular attack of the hemiaminal carbon by the enamine carbon followed by a new nucleophilic rearrangement to form tetrahydropyrimidine cycle; the forth step was closure of the lactam ring, if ester group was available as in trifluoropyruvate.2013 Elsevier B.V. All rights reserved.

Full text of DOI:10.1016/j.jfluchem.2013.10.017

Journal of Fluorine Chemistry 157 (2014) 1–11
Contents lists available at ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
New nucleophilic rearrangement in the mechanism of the
three-component domino cyclisation affording fluoroalkylated
(pyrrolo)quinazolines
a,
b
a
a
*
ˇ
´
ˇ´
ˇ
´ˇ
Oldrich Paleta , Bohumil Dolensky , Jirı Palecek , Jaroslav Kvıcala
a
´
Department of Organic Chemistry, Institute of Chemical Technology in Prague, Technicka 5, 166 28 Prague 6, Czech Republic
b
´
Department of Analytical Chemistry, Faculty of Chemical Engineering, Prague Institute of Chemical Technology, 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:
The following mechanism steps were verified for the three-component domino cyclisation affording
(pyrrolo)quinazolines from 2-(aminomethyl)aniline, a very reactive oxo compound and ‘‘usual’’ oxo
compound. The first step was a rapid reaction of very reactive oxo compound (trifluoropyruvate or
hexafluoroacetone) with benzylic amino group to form hemiaminal, but not imine; the second step was
the reaction of oxo compound with aromatic amino group to form imine (Schiff base) being in
equilibrium with its enamine form; the third step was an intramolecular attack of the hemiaminal
carbon by the enamine carbon followed by a new nucleophilic rearrangement to form tetrahydropyr-
imidine cycle; the forth step was closure of the lactam ring, if ester group was available as in
trifluoropyruvate.
Received 2 September 2013
Received in revised form 22 October 2013
Accepted 25 October 2013
Available online 5 November 2013
Keywords:
Three-component cyclisation
Mechanism
Substituted pyrroloquinazolines
Trifluoropyruvate aldols
Enamine nucleophiles
Nucleophilic rearrangement
ß 2013 Elsevier B.V. All rights reserved.
1. Introduction
compounds in mixtures of linear (5) and angular products
(6, Scheme 1) and their stereoisomers, a NMR method has been
Recently, we have published a study of diastereoselectivity of
the three-component domino cyclisation involving 2-(amino-
methyl)aniline (1), methyl 3,3,3-trifluoropyruvate (2) and ketone
or aldehyde (3, Scheme 1) [1]. This cyclisation afforded trifluor-
omethylated pyrroloquinazolines (5 and 6). The major products 5,
formed by the linear annulation of the lactam ring were obtained in
relative yields usually above 85% [1]. The minor products 6 possess
angularly annulated lactam ring. The complete cyclisation regios-
electivity in exclusive formation of linear (5) or angular products
(6) was obtained when the starting N-monoalkylated diamine 1
was employed [1].
The cyclisations in Scheme 1 belong to multicomponent
reactions (MCRs) [2–5], which have a great advantage in saving
laboratory work: mixing together three or more components leads
to their stepwise chemical combining by a characteristic chemical
algorithm to form an end product. New [6], novel [7] or novelized
[8] MCRs have appeared in recent literature.
elaborated [9]. By using this method, a facile assignment of
the relative configuration on stereogenic centers using the
combination of homo- (proton–proton) and heteronuclear (pro-
ton–fluorine) NOE experiments was possible. This approach also
allowed to assign the relative configuration in minor products
down to 1% of their relative content [9]. We have applied this
methodology for structure elucidations in this paper.
2. Results and discussion
The cyclisation resembles the Mannich synthesis [5] involving
an amine and two oxo compounds, one of which being more
reactive than the second one. The higher reactive oxo compound
starts the Mannich domino sequence by the reaction with the
amine. In our cyclisation [1], the beginning of the domino sequence
is similar (vide infra).
In this paper we would like to report the results of the study of
the reaction mechanism of the introduced three-component
domino cyclisation. For the structure elucidation of the individual
2.1. Closure of the lactam ring as the last step
The closure of the lactam ring in the intermediate 4 to form the
final pyrroloquinazoline products 5 and 6 (Scheme 1) should
proceed as relatively rapid reaction, because we have never
detected patterns of the intermediate 4 during the cyclisation or in
final reaction mixtures.
*
Corresponding author. Tel.: +420 22044 4288.
E-mail address: oldrich.paleta@vscht.cz (O. Paleta).
0022-1139/$ – see front matter ß 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jfluchem.2013.10.017

2
O. Paleta et al. / Journal of Fluorine Chemistry 157 (2014) 1–11
COOCH3 R1
Scheme 3). On the other hand, the intermediate 4a, which is an
F3C
R2
NH2
NH2
aminal, could also be formed potentially from the aldol 11 by its
reaction with the diamine 1 (pathway B, Scheme 3). Similar
cyclisations involving oxoesters have been reported [11]. Our
reaction of the aldol 11 with diamine 1 afforded the products 13
and 16 in isolated yields of 31% and 1%, respectively. Despite of a
different character of the pathways A and B, the cyclisation
regioselectivity as well as the diastereoselectivity of the cyclisa-
tions were close (Table 1). The diastereoisomers 13a–b and 16a–b
are denoted cis/trans relatively to the hydroxy group. The cis-
configurations, in which bulky phenyl substituents are in the trans-
position relatively to the hydroxy group (14a–16a), were highly
prevailing in linear products. The portion of the angular cyclisation
(products 16–18) was very low in the range of 2–6.5% rel. The
aldols 11 or 12 were formed as by-products during the pathway A.
A higher temperature supressed the amount of the aldol 11 to zero
yield.
+
+
2
3
O
O
-H2O
1
R2
R1
H
N
CF3
OH
COOCH3
NH
4
OH
R2
O
N
CF3
R2
R1
H
N
CF3
R1
N
OH
NH
O
5
6
major product
minor product
Scheme 1. General scheme of the three-component cyclisation.
For a qualified estimation of the contributions of the individual
pathways A and B to the formation of cyclisation products, we
followed the kinetics of the reactions 10 + 3a (path A) and 11 + 1
(path B). The graphic results in Fig. 1 revealed that the cyclisation
of the hemiaminal 10 (curve A) was much faster than the
cyclisation of the aldol 11 (curve B).
The different kinetics also indicated different mechanisms for
the pathways A and B. The conclusion has been that the pathway B
could contribute to the yields of individual products not more than
5–7% rel. (at room temperature), which was insignificant for the
cyclisation result.
To verify that the closure of the lactam ring is not a necessary
step determining the occurance of the three-component cyclisa-
tion (Scheme 2), we performed the following experiments: the
first, in the three-component mixture, trifluoropyruvate (2) was
substituted with hexafluoroacetone (7), which afforded with
diamine 1 and ketone 3a substituted quinazoline derivative 8 as
the result; the second, the hemiaminal 9 was prepared separately,
to which acetophenone (3a) was added to form again the product
of the three-component cyclisation (8). It can be concluded that the
closure of the lactam ring is not a factor determining the three-
component cyclisation.
Ar
CH3
3a-c
+
+
1a
2a
O
2.2. Hemiaminal of trifluoropyruvate (10) as the first intermediate
We have published recently [10] that the hemiaminal 10 was
formed immediately after mixing trifluoropyruvate 2 with diamine
1. Hemiaminal 10 then reacted [1] with ketones added (3a–c) to
form the corresponding reactive intermediates 4a–c (pathway A,
NH2
Ar
OH
CF3
OCH3
H
CF3
OH
N
O
O
10
COOCH3
11-12
Ar
O
NH2
-H2O
-CH3OH
-H2O
-CH3OH
CH3
NH2
3a-c
1
F3C
CF3
CH3
Ar
1 +
+
H
N
B
A
O
CF3
OH
3a
7
O
NH
O
OCH3
4a-c
Y
H
N
CF3
OH
CF3
OH
O
N
NH
H
CF3
N
OH
8 (58%)
NH
CF3
N
3a
Y
O
16-18
13-15
NH2
H
N
1
7
CF3
OH
+
3a, 4a, 11, Ar = C6H5; 3b, 4b, 12, Ar = 4-NO2C6H4;
3c, 4c, Ar = 4-CH3OC6H4.
13, 16, Y = H; 14, 17, Y = NO2; 15, 18, Y = OCH3
9
CF3
Scheme 3. Two possible mechanistic pathways A and B leading to the common
reactive intermediates 4a–c in the cyclisation of acetophenones.
Scheme 2. Hexafluoroacetone (7) as the reactive oxo compound for the three-
component cyclisation.

O. Paleta et al. / Journal of Fluorine Chemistry 157 (2014) 1–11
3
Table 1
Stereochemistry of the products 13–15 and 16–18 formed by the pathways A or B
(Scheme 3) at 100 8C.
Y
Y
H
N
H
N
OH
OH
CF3
CF3
N
N
cis-
13a-15a
O
trans-
O
13b-15b
CF3
OH
CF3
O
O
OH
N
N
NH
NH
cis-
16a-18a
Y
Y
trans-
16b-18b
Fig. 1. Kinetics of the cyclisation reactions A and B (Scheme 3): curve A: reaction
pathway A (components 10 + 3a); curve B: reaction pathway B (components 11 + 1).
Pathway
Y
Products (%rel.)
13a
13b
16a
16b
Aldol (%)
diamine 1 afforded only the quinazoline 20 in 38% yield, a cyclic
aminal of the pyruvate 19 as the sole product (Scheme 4). When
methyl pyruvate (19) was employed with 2-(aminomethyl)aniline
(1) and cyclopentanone (3d), then a mixture of products was
obtained as follows: the quinazoline derivative 20 (45%) as the
main product and only very small amount of the product of the
three-component cyclisation (21, 3%) and its dehydrated deriva-
tive 22 (2%).
In contrast, methyl pyruvate (19) could act as an ‘‘usual’’ oxo
compound when used with trifluoropyruvate (2) and diamine 1:
This starting mixture afforded the main compound 23 (58%) as the
product of the three-component reaction, in which pyruvate (19)
acted as the less reactive oxo compound; the second product was
the aminal – quinazoline 20, while unreacted trifluoropyruvate
was converted to its hydrate 24.
A
H
93a
94
2
3
0
0
2
0
3
39/rt, (11)
0/100 8C
6
B
H
Y
84
13
Pathway
Products (%rel.)
14a
14b
17a
17b
A
NO2
Y
86.5b
7
6.5
0
7/100 8C (12)
0/100 8C
Pathway
Products (%rel.)
15a
15b
18a
ꢀ2
18b
A
OMe
ꢀ98c
0
0
The cyclisation regioselectivity linear vs. angular (%rel.) was as follows:
a
13:16 = 95:5.
b
14:17 = 93.5:6.5.
c
15:18 ffi 98:2.
Trifluoropyruvate (2) as well as hexafluoroacetone (7) do not
form aminals and do not react with secondary amines to form the
corresponding hemiaminals [10,12]. This is the fundamental
difference between trifluoropyruvate 2 and pyruvate 19. It can
be concluded in this section that the three-component cyclisation
occurs with the diamine 1, if (a) a very reactive oxo compound (e.g.
trifluoropyruvate 2 or hexafluoroacetone 7) forming hemiaminals,
not forming aminals and not reacting with secondary amines is
involved; (b) a second oxo compound exhibiting usual properties is
present in the reaction mixture.
2.3. Formation of the pyrimidine ring in the three-component
cyclisation
2.3.1. Crucial role of 3,3,3-trifluoropyruvate (2) or hexafluoroacetone
(7) in the three-component cyclisation
The crucial role of a highly reactive oxo compound in the
cyclisation was verified by the experiments involving methyl
pyruvate (19) as depicted in Scheme 4: the reaction of 19 with the
COOCH3
H3C
H
N
CH3
COOCH3
19
O
NH
20 (38%)
COOCH3
H3C
H
N
H
N
H
N
CH3
COOCH3
+
NH2
NH2
19
3d
CH3
OH
O
O
+
+
N
NH
20 (45%)
N
O
O
21 (3%)
22 (2%)
1
COOCH3
F3C
COOCH3
H3C
COOCH3
H
H
CH3
COOCH3
+
2
N
19
N
O
F3C
+
O
CF3
OH
COOCH3
+
NH
20 (41%)
N
OH
HO
24
23 58%)
cis/trans 54/46
O
Scheme 4. Cyclization reactions involving methyl pyruvate.

4
O. Paleta et al. / Journal of Fluorine Chemistry 157 (2014) 1–11
CH2
NH
Ph
CH3
Ph
+ PhCOCH3
NH2
N
3a
CF3
H
N
H
H
N
CF3
OH
COOCH3
CF3
- H2O
N
26
HO
COOCH3
COOCH3
HO
25
10
Ph
H
N
H
N
Ph
CF3
OH
OCH3
OH
CF3
NH
N
H
O
COOCH3
4a
26
concerted or stepwise
Scheme 5. Formation of the reactive enamine intermediate 26 and its ring closure to form the pyrimidine ring in 4a by a new nucleophilic rearrangement process similar to a
(2 + 2)-metathetic rearrangement.
2.3.2. Evidence of the intramolecular enamine attack on the
hemiaminal carbon and new nucleophilic rearrangement
which the lactam ring was closed to form E. This intermediate was
subsequently stabilized by the elimination of piperidine to form
37. The sequence 33 + 36 ! D ! E is analogous to the mechanism,
by which the cyclisation depicted in Schemes 1 and 7 occurs. In the
case of the enamines 32 or 34–35, the reaction with the
hemiaminal 36 led to less stable intermediates F, which obviously
eliminated piperidine before lactam ring closure to afford Schiff
bases 38–40. These compounds were hydrolysed during purifica-
tion on silica gel to the final aldols 41–43.
The conversion of the hemiaminal 10 to the intermediate 4a
includes several mechanistic (probably equilibrium) steps
(Scheme 5). During this process, the hemiaminal C–N bond in
10 is probably detached and the hemiaminal (marked) carbon
combined with the methyl of acetophenone (3a). This transfor-
mation can be understood as a kind of aldol reaction between 3a
and trifluoropyruvate 2. It has recently been verified that enamines
react easily with trifluoropyruvate 2 to form aldol products [13].
An analogous intramolecular reaction leading to the quinazoline
intermediate 4a could proceed via the enamine 26. The following
process can be proposed for such transformation: the Schiff base
25 is in an equilibrium with the enamine tautomer 26, in which the
highly nucleophilic carbon then attacks the hemiaminal carbon
under the cleavage of the hemiaminal C–N bond and closure of
tetrahydropyrimidine cycle. This intramolecular transformation
leading to the intermediate 4a can be a concerted
Exclusive role of the hemiaminal 10 in the cyclisation could be
underlined by the fact that 2-aminobenzamide (27a, Scheme 6) or
2-(methylamino)benzamide (27b), both bearing amide groups of
low basicity, did not form hemiaminals with trifluoropyruvate (2),
nor the three-component cyclisation occurred. Instead, aldol 28
was formed from unreacted pyruvate 2 and cyclopentanone (3d).
However, when the aldolisation was blocked by using dimethylk-
etal 29 instead of free acetone, then regiospecific angular
cyclisation occurred at the more basic amino group with low
yield of 30. During the course of the cyclisation, benzamide 27a
probably formed first Schiff base A, which was in equilibrium with
highly unstable hemiaminal B that rearranged to the intermediate
C. Similarly, hemiaminal 10 reacted with 29 by regiospecific
linear closure of the lactam ring at the more basic nitrogen atom
to afford 31 [9].
We have verified that the proposed enamine reaction
depicted in Scheme 5 is possible. The model reaction was arranged
in such manner that the series of enamines 32–35 prepared
from piperidine was reacted with benzylamine hemiaminal 36
(Scheme 7) by an intermolecular cyclisation, which is usually 100–
1000 times slower then the corresponding intramolecular reaction.
However, the reactions proceeded smothly and were finished in
several hours, while the three-component cyclisations proceeded
several days at r.t. The products of the reactions were dependent on
the enamine structure. The reaction of the enamine derived from
cyclohexanone (33) was the most informative by affording the
indolone derivative 37 as the final product in high yield. It can be
deduced from its structure that the aminal D was formed first, in
NH-R
OH
CF3
+ 2
+
NH2
O
COOCH3
28
O
O
27a-b
27a
(ca. 60%)
58%)
27a, R = H; 27b, R = CH3
HO
CF3
CH3
O
NH2
H3C
CH3
+
+
2a
NH2
N
CH3O OCH3
O
29
NH
> 25%
O
30
CH3
N
H
N
NH2
CF3
OH
+
29
H
N
CF3
O
29%
31
10
HO COOCH3
(CH3)2C(OCH3)2
+
27a
CH3
CH3
CH3
CH3
N
N
2
B
H
N
CF3
NH2
COOCH3
OH
O
O
A
CH3
H
N
CF3
OH
COOCH3
cyclisation
31
NH
O
C
Scheme 6. Cyclisations of 2-(methyl)aminobenzamides 27a–b and hemiaminal 10
with 2,2-dimethoxypropane (29).

O. Paleta et al. / Journal of Fluorine Chemistry 157 (2014) 1–11
5
H
R1
CF3
CF3
OH
CF3
OH
OH
CF3
OH
COOMe
R2
C
C
H
OMe
O
33
O
N
N
H
N
N
H
N
O
N
N
37
(75%)
D
36
E
32-35
CF3
C
H
CF
1
3 OH
R1
R2
H
R1
O
OH
OMe
CF3
R
C
2
R
OH
COOMe
41-43
R2
OMe
N
O
N
O
N
H
38-40
(51-75%)
F
32, 38, 41, R1-R2 = -(CH2)3-; 33, 37, R1-R2 = -(CH2)4-; 34, 39, 42, R1 = H, R2 = Ph;
35, 40, 43, R1 = H, R2 = CH3
Scheme 7. Evidence of the intramolecular enamine attack in the course of the three-component cyclisation.
2.3.3. Stereochemistry of the cyclisation
In which reaction step or steps is the stereochemistry of the
products arranged? The last step, lactamisation of the structure 4,
proceeds as a SN reaction at the carbonyl group by the exo-trig
process [14] without forming
a stereogenic carbon (4a–4c,
Schemes 3 and 5). On the other hand, racemic hemiaminal 10 is
formed in the first step, but no new stereogenic centers are created
in the subsequently formed intermediates 25 and 26 (Scheme 5).
Thus, the stereochemistry of the final products (e.g. 13, Scheme 8)
should be created in the rearrangement step during the
tranformation of the intermediate 26 into intermediate 4a. As
an example, the enantiomer (R)-26 of the intermediate 26 should
be transformed into the intermediates cis- and trans-4a affording
the final mixture of cis-13a and trans-13 (Table 1). An analogous
stereochemical scheme can be drawn for the cyclisation of the
enantiomer (S)-26 and similar schemes for the cyclisations
involving an aldehyde or cyclic ketone.
2.3.4. The effect of para-substituents in acetophenones on the rate of
the cyclisation
We compared the effect of the para-nitro (3b) and para-
methoxy (3b) groups in acetophenone (3a) on the kinetics of the
three-component cyclisation. The kinetic dependences are shown
in Fig. 2. Non-substituted acetophenone appeared to be the most
Ph
H
N
H
N
OH
OH
Ph
H
N
H
N
CF3
CF3
H3COOC
H3COOC
(R)-26
(R)-26
cis-4a
H
N
Ph
H
Ph
CF3
OH
CF3
OH
N
NH
O
NH
OCH3
OCH3
O
trans-4a
trans-13a
(R,S)-13a
Fig. 2. Effect of para-substituents in acetophenone parent structure on the over-all
kinetics of the three-component cyclization involving 1a and 2a (Scheme 3): curve
A: acetophenone (3a); curve B: p-nitroaceto-phenone (3b); curve C: p-
methoxyacetophenone (3c).
cis-13a
(R,R)-13a
Scheme 8. Stereochemistry of the nucleophilic rearrangement.

6
O. Paleta et al. / Journal of Fluorine Chemistry 157 (2014) 1–11
exhibiting usual properties is present in the reaction mixture.
CH3
Both p-electron-donating or p-electron-withdrawing groups in
acetophenone retarded the rate of the cyclisation. It was verified
that the cyclisation did not occur as a four-component reaction.
CF3
OH
COOCH3
O
H3C
NH2
CH3
O
O
NH2
F3C
43
4. Experimental
COOCH3
2
4.1. General
H
N
CH3
See section Supplementary data.
4.2. Preparation of the starting compounds
Ph
Ph
CF3
OH
N
O
44
Methyl 3,3,3-trifluoropyruvate (2) was synthesized by the
reaction of hexafluoropropene oxide with sulfuric acid according
to our procedure [16]. Hemiaminal 10 (methyl 2-(2-aminobenzy-
l)amino-3,3,3-trifluoro-2-hydroxypropanoate) was synthesized
according to our procedure [10]. Benzamides 27a and 27b were
prepared by modified procedures [17a,17b,18]. Aldol of hexa-
fluoroacetone with acetophenone (11) was prepared according to
Ref. [19]. Modified procedures were used to prepare 1-(cyclopent-
1-en-1-yl)piperidine (32) [20], 1-(cyclohex-1-en-1-yl)piperidine
(33) [21], 1-(1-phenylethen-1-yl)piperidine (34) [22] and (prop-1-
en-2-yl)piperidine (35) [23] (see Supplementary data). Aldol 11
(methyl 2-hydroxy-4-(4-nitrophenyl)-4-oxo-2-(trifluoromethyl)-
butanoate) was prepared according literature [24].
Scheme 9. Attempt to realize a four-component domino cyclization.
reactive, while electron-withdrawing nitro group (Hammett
p = +81) [15a] or electron-donating methoxy group (Hammett
s
s
p = À0.28) [15a] decreased the overall rate of the cyclisation.
Probably, the nitro group decreases the nucleophilicity of the
enamine 26 (Scheme 5), while the methoxy group, similarly as
methyl group (Hammett
s
p = À0.14) [15a] destabilizing enamines
[15b,c] can shift the imine 25–enamine 26 equilibrium to the left
side, decreasing thus the concentration of the enamine 27. The
experimental data have shown that a negative effect of the
methoxy group on the cyclisation rate appeared to be stronger
than that of the nitro group (Fig. 2).
4.3. 1,1,1,3,3,3-Hexafluoro-2-[(2-phenyl-1,2,3,4-
Regioselectivity and diastereoselectivity of the three-compo-
nent cyclisation involving acetophenones 3a–c (Scheme 3) are
summarized in Table 1 and demostrate high selectivity of the
cyclisation: linear annulation yielded the products 13–15 in the
range of 94–98% rel., among which the cis-diastereoisomers
dominated in the range of 86–98% rel.
tetrahydroquinazolin-2-yl)methyl]propan-2-ol (8)
4.3.1. 2-[2-(Amino)benzylamino]-1,1,1,3,3,3-hexafluorpropan-2-ol
(9)
A flask (250 mL) was charged with dry CH2Cl2 (100 mL) and
diamine 1 (0.30 g, 2.46 mmol) and gaseous hexafluoroacetone
(0.41 g, 2.46 mmol) was introduced inside the liquid at rt while
stirring. The reaction was finished in 5 min (check by GC and 19F
NMR). Volatile components were then removed in vacuum (rotary
evaporator) to afford 0.67 g (94%) of hemiaminal 9 as honey-like
material.
2.4. Can the cyclization occur as a four-component reaction?
The cyclisation reaction between the enamine 33 and
heminal 36 affording the product 37 (Scheme 7) can be looked
as a four-component cyclisation, because each of the reactants
had been prepared from two components. A question had arosen
whether the cyclisation could occur as a real four-component
reaction. In the experiment (Scheme 9), 2-(aminomethyl)-
aniline (1) was substituted with aniline and benzylamine, while
trifluoropyruvate (2) and acetone were applied as usual oxo
compounds. As a result, the expected cyclisation product 44 was
not detected in the reaction mixture, only aldolisation occurred
between acetone and trifluoropyruvate to afford the aldol 43
(Scheme 9).
1H NMR (300 MHz, CDCl3):
d 2.21 (bs, 1H, OH), 4.03 (s, 2H,
CH2), 5.72 (bs, 3H, NH and NH2), 6.78 (d, 1H, Ph, J = 7.7 Hz), 6.92
(t, 1H, Ph, J = 7.4 Hz), 7.18 (m, 2H, Ph) ppm. 13C NMR (75.5 MHz,
CDCl3):
J = 291 Hz), 118.7, 121.4, 125.0, 129.3, 130.7, 143.1 (Ph) ppm.19
NMR (376.6 MHz, CDCl3):
d 43.4 (CH2), 84.1 (q, C–CF3, J = 30.4 Hz), 122.1 (q, CF3,
F
d
À79.1 (s, CF3) ppm. EIMS: m/z (rel.
int.): 288 (88) [M]+, 272 (4), 241 (4), 240 (16), 219 (68), 207 (44),
202 (100), 200 (24), 194 (12), 169 (28), 149 (15), 147 (15), 143
(28), 132 (40), 121 (36), 120 (40), 119 (40), 107 (28), 106 (20),
105 (15), 103 (32), 94 (15), 93 (28), 92 (24), 91 (12), 83 (16), 76
(37), 70 (28), 69 (16), 57 (20), 55 (20), 50 (15), 42 (15). For X-ray
structure see Supplementary data, p. 5.
3. Conclusion
It has been verified and proved that the three-component
cyclisation includes the following subsequent steps: (1) Formation
of hemiaminal 10 (or 9) of trifluoropyruvate 2 (or hexafluoroa-
cetone 7) as the first intermediate (Scheme 3); (2) formation of a
Schiff base 25 on the aniline amino group in diamine 1 (Scheme 5);
(3) tautomeric equilibrium between the Schiff base 25 and its
enamine tautomer 26 (Scheme 5); (4) intramolecular enamine
attack on the hemiaminal carbon to form quinazoline ring 4a
(Scheme 5); (5) closure of the lactame ring as not a necessary step if
ester group is not available (product 8, Scheme 2). The three-
component cyclisation occurs with diamine 1, if: (a) a very reactive
oxo compound (e.g. trifluoropyruvate 2 or hexafluoroacetone 7)
forming hemiaminal, not forming aminal and not reacting with a
secondary amine, is involved; (b) a second oxo compound
4.3.2. Reaction of hexafluoroacetone hemiaminal 9 with
acetophenone: products 8 and 11
A solution of acetophenone (0.3 g, 2.46 mmol) in CH2Cl2
(20 mL) was added dropwise to a solution of hemiaminal 9 in
40 mL of CH2Cl2 prepared from diamine 1 (0.3 g, 2.46 mmol) and
hexafluoroacetone (7, 0.41 g, 2.46 mmol), and the cyclisation was
monitored by 19F NMR at rt for 60 days. Solvent was evaporated
and the residue purified by column chromatography (SiO2 40 g,
CH2Cl2/hexane 4:1): Fraction 1 was aldol 11 (0.15 g, 30.7%), mp
32–33 8C (Ref. [18] mp 33–34 8C). Fraction 2 was the product 8
(0.55 g, 57.4%), which was crystallized (CHCl3/hexane 1:1) to get
mp 103–106 8C.
1,1,1,3,3,3-Hexafluoro-2-[(2-phenyl-1,2,3,4-tetrahydroquina-
zolin-2-yl)methyl]propan-2-ol (8)

O. Paleta et al. / Journal of Fluorine Chemistry 157 (2014) 1–11
7
1H NMR (300 MHz, CDCl3):
d
1.55 (bs, 1H, OH), 2.31 (d, 1H, CH2,
4.6.1. (2R*,3aS*)-2-Hydroxy-3a-(4-nitrophenyl)-2-
(trifluoromethyl)-3,3a,4,9-tetrahydropyrrolo-[2,1-b]quinazolin-
1(2H)-one (14c)
J = 16.5 Hz), 2.49 (d, 1H, CH2, J = 16.5 Hz), 2.92 (bs, 1H, NH), 3.64 (d,
1H, CH2, J = 16.2 Hz), 3.88 (d, 1H, CH2, J = 16 Hz), 4.62 (bs, 1H, NH),
6.69–6.85 (m, 3H, Ph-H), 7.12 (t, 1H, Ph-H, J = 7.4 Hz), 7.24–7.52
(m, 5H, Ph-H) ppm. 13C NMR (75.5 MHz, CDCl3):
41.78 (CH2), 124.29 (q, CF3, J = 281 Hz), 116.8, 120.1, 125.9, 126.4,
127.6, 128.3, 128.8, 129.2, 139.8, 142.4 (2Â Ph) ppm. Signals of C–
CH2 and C–CF3 were not assigned due to weak intensity. 19F NMR
Crystals obtained above were pulverized and dried (60 8C,
13 Pa, 4 h), mp above 204 8C. IR: 1715 (s), 1694 (s) cmÀ1 1H NMR
(300 MHz, CD3COCD3): 8.17 (d, 2H, J = 8.8 Hz), 7.64 (d, 2H,
d
37.76 (CH2),
.
d
J = 8.8 Hz), 7.14–7.06 (m, 1H), 6.89 (d, 1H, J = 7.7 Hz), 6.86–6.80 (m,
2H), 5.29 (s, 1H), 5.02 (d, 1H, J = 17 Hz), 4.96 (bs, 1H), 3.9 (d, 1H,
J = 17 Hz), 2.83 (d, 1H, J = 15.5 Hz), 2.59 (d, 1H, J = 15.5 Hz) ppm.
(376.6 MHz, CDCl3):
Supporting info. C18H16F6N2O (319.12): calcd C 55.39, H 4.13, N
7.18; found C 55.68, H 4.38, N 7.12.
d
À78.5 (s, CF3) ppm. For X-ray analysis see
13C NMR (75 MHz, CD3COCD3):
d
169.2, 149.2, 143.7, 140.3, 128.8,
128.2 (2C), 127.5, 125.1 (2C); 124.2 (qa, J = 284 Hz), 77.2 (qa,
J = 31.7 Hz), 74.6, 44.8, 40.5 ppm. 19F NMR (75 MHz, CDCl3):
À80.2 (s) ppm. NOEDif: see Supplementary information, p. 7.
18H14F3N3O4 (393,32): calcd C 54.97, H 3.59, N 10.68; found C
4,4,4-Trifluoro-3-hydroxy-1-phenyl-3-(trifluoromethyl)butan-1-
d
one (11). 1H NMR (300 MHz, CDCl3):
d
3.48 (s, 2H, CH2), 7.29 (bs,
1H, OH), 7.54 (t, 2H, Ph-H, J = 7.9 Hz), 7.69 (t, 1H, Ph-H, J = 7.4 Hz),
7.98(d, 2H, Ph-H,J = 7.9 Hz) ppm. 13CNMR(75.5 MHz, CDCl3):
31.9
C
d
54.93, H 3.69, N 10.66.
(CH2), 76.8 (q, C–CF3, J = 30.1 Hz), 122.7 (q, CF3, J = 288.3 Hz), 128.5,
129.1, 135.2, 135.5 (Ph); 199.7 (C55O) ppm. 19F NMR (376.6 MHz,
4.6.2. (2R*,3aR*)-2-Hydroxy-3a-(4-nitrophenyl)-2-
(trifluoromethyl)-3,3a,4,9-tetrahydropyrrolo[2,1-b]-quinazolin-
1(2H)-one (14t)
CDCl3):
d
À78.5 (s, CF3) ppm. EIMS: m/z (rel. int.): 286 (3) [M]+, 217
(2), 106 (12), 105 (100), 97 (2), 78 (7), 77 (57), 69 (7), 51 (18).
1H NMR (300 MHz, CDCl3):
J = 17 Hz) ppm. 19F NMR (75 MHz, CDCl3):
d
4.92 (d, 1H, J = 17 Hz), 3.88 (d, 1H,
À79.8 (s) ppm.
4.3.3. Three-component cyclisation of 2-(aminomethyl)aniline (1)
with hexafluoroacetone (7) and acetophenone (3a): product 8
A flask (250 mL) was charged with dry CH2Cl2 (80 mL), diamine
1 (0.30 g, 2.46 mmol) and acetophenone (3a, 0.30 g, 2.46 mmol)
and gaseous hexafluoroacetone (7, 0.41 g, 2.46 mmol) was
introduced immediately inside the liquid at rt while stirring,
which was continued for 12 days at rt. Volatile components were
removed at rt in vacuum and the residue was purified by CC (SiO2,
CH2Cl2/hexane 4:1). The second fraction contained 8, which after
evaporation afforded pure 8 (0.275 g, 28.7%). For analytical data
see the preceding experiment.
d
4.6.3. (2R*,3aS*)-2-Hydroxy-3a-(4-nitrophenyl)-2-(trifluoromethyl)-
3,3a,4,5-tetrahydropyrrolo[1,2-a]-quinazolin-1(2H)-one (17c)
1H NMR (300 MHz, CD3COCD3):
d 8.59 (d, 1H, J = 8.2 Hz), 8.19
(d, 2H, J = 8.8 Hz), 7.73 (d, 2H, J = 8.8 Hz), 7.31 (t, 1H, J = 8 Hz), 7.1
(dt, 1H, J = 1.1 and 7.7 Hz), 7.02 (d, 1H, J = 7.7 Hz), 6.28 (s, 1H), 4.02
(d, 1H, J = 17.6 Hz), 3.57 (d, 1H, J = 17.6), 2.83 (bs, 1H), 2.77 (1H,
J = 14.3 Hz), 2.67 (1H, J = 14.3) ppm. 13C NMR (75 MHz,
CD3COCD3):
d
166.9, 150.9, 149.3, 136.0, 130.1 (2 Â CH),
128.7 (CH), 127.6 (CH), 127.4, 126.2 (CH), 125.6 (qa, J = 284 Hz),
125.1 (2 Â CH), 121.2 (CH), 78.7, 77.7 (qa, J = 31 Hz), 46.0,
44.0 ppm. 19F NMR (75 MHz, CDCl3):
d
À80.1 (s) ppm.
4.4. General procedures for the three-component cyclisations of 2-
(aminomethyl)aniline (1), methyl trifluoropyruvate (2) and oxo
compound
C18H14F3N3O4 (393,32): calcd C 54.97, H 3.59, N 10.68; found C
54.33, H 4.12, N 10.44.
Detailed procedures together with analytical data are given in
Supplementary data. All reactions were performed in an oven-
dried apparatus. Reaction flask was usually equipped with reflux
condenser, to which a drying tube was attached (CaCl2/KOH).
Typical procedure A: A solution of methyl 3,3,3-trifluoropyr-
uvate (2, 315 mg, 2 mmol) in dioxane (10 mL) was added dropwise
to a solution of 2-(aminomethyl)aniline (1, 245 mg, 2 mmol) in
dioxane (10 mL) at rt while stirring. After 5 min, a solution of an
oxo compound (3, 2 mmol) in dioxane (6 mL) was added dropwise
to the mixture at rt while stirring, which was continued until the
end of the reaction (100–200 h, check by NMR). Volatile
components were removed by rotary evaporator and the crude
product was separated to individual cyclisation products by
column chromatography (CC) on silica gel. For analytical purposes,
some products were purified further by crystallization.
4.6.4. Methyl 2-hydroxy-4-(4-nitrophenyl)-4-oxo-2-
(trifluoromethyl)butanoate (12)
1H NMR (300 MHz, CDCl3):
d
8.34 (d, 2H, J = 8.8 Hz), 8.1 (d, 2H,
J = 8.8 Hz), 4.21 (bs, 1H), 3.93 (s, 3H), 3.81 (d, 1H, J = 17.6 Hz), 3.71
(d, 1H, J = 17.6 Hz) ppm. 13C NMR (75.5 MHz, CDCl3):
193.3, 169,
150.8, 140.1, 129.3, 124, 123 (qa; J = 286.3 Hz), ca. 75 (qa, overlap
with CDCl3), 54.4, 41.1 ppm. 19F NMR (75 MHz, CDCl3):
d
d
À79.2
(s) ppm. C12H10F3NO6 (321.2): calcd C 44.87, H 3.14, N 4.36; found
C 44.95, H 3.40, N 4.26.
4.7. Cyclisation involving p-methoxyacetophenone (3c): products 15c
and 18c
For experimental details see Supplementary data.
Typical procedure B: The same amounts of the reactants 1, 2, 3
and solvent as in Procedure A. After mixing 1 and 2, the mixture
was heated up to 100 8C and an oxo compound 3 was added
dropwise and the mixture stirred for 1–2 h (end of the reaction).
The resulting reaction mixture was cooled to rt and then treated as
in Procedure A.
4.7.1. (2R*,3aS*)-2-Hydroxy-3a-(4-methoxyphenyl)-2-
(trifluoromethyl)-3,3a,4,9-tetrahydro-pyrrolo[2,1-b]quinazolin-
1(2H)-one (15c)
IR: 1697 (s) cmÀ1 1H NMR (300 MHz, CDCl3):
. d 7.32 (d, 2H,
J = 8.8 Hz), 7.1 (t, 1H, J = 7.4 Hz), 6.90–6.78 (m, 3H), 6.85 (d, 2H,
J = 8.8 Hz), 5.02 (d, 1H, J = 16.5 Hz), 4.55 (bs, 1H), 3.9 (d, 1H,
J = 17 Hz), 3.76 (s, 3H), 3.49 (bs, 1H), 2.69 (d, 1H, J = 15.4 Hz), 2.5 (d,
4.5. Cyclization involving acetophenone (3a): products 13 and 16
For analytical data of 13c, 13t, 16c and 16t see Ref. [1]
1H, J = 15.4 Hz) ppm. 13C NMR (75 MHz, CDCl3):
d 168.4 (C), 159.6
(C), 140.6 (C), 132.9 (C), 127.8 (CH), 127.5 (2CH), 126.7 (CH), 123.7
(C), 120.8 (CH), 118.7 (C), 118.4 (CH), 114.5 (2CH), 76.8 (qa, C,
J = 31.5 Hz), 73.9 (C), 55.3 (CH3), 44.8 (CH2), 39.8 (CH2) ppm. 19F
4.6. Cyclisation involving p-nitroacetophenone (3b):
products 12, 14 and 17
NMR (75 MHz, CDCl3):
calcd C 60.32, H 4.53, N 7.40; found C 59.97, H 4.55, N 7.36. NOEDif:
see Supplementary data, p. 8.
d
À80.4 (s) ppm. C19H17F3N2O3 (378.35):
For experimental details see Supplementary data.

8
O. Paleta et al. / Journal of Fluorine Chemistry 157 (2014) 1–11
4.7.2. 2-Hydroxy-3a-4-(methoxyphenyl)-2-(trifluoromethyl)-
dioxane (15 mL) was heated to 100 8C for 12 days while stirring.
After evaporation of volatile components, raw oily material
(1.12 g) was obtained, which was separated by CC (SiO2, CH2Cl2,
then CHCl3/Et2O 3:2) in two fractions: fraction 1 (552 mg, 57.9%)
was a mixture of two stereoisomers (23c: 23t = 54:46), which by
repeated crystallization (1x CHCl3, 2x CH2Cl2) afforded pure 23c;
fraction 2 was pure quinazoline 20 (243 mg, 40.8%)–for data
see 4.9.1.
3,3a,4,5-tetrahydropyrrolo[1,2-a]-quinazolin-1(2H)-one (18c)
1H NMR (300 MHz, CDCl3):
d 8.54 (d, 1H, J = 8.2 Hz) ppm
4.8. Reaction of aldol 11 with diamine 1: products 13 and 16
For product characterization see Section 4.5.
4.9. Cyclisation involving methyl pyruvate (19) and cyclopentanone:
products 20, 21 and 22
4.11.1. (2R*,3aR*)-Methyl-2-hydroxy-1-oxo-2-(trifluoromethyl)-
1,2,3,3a,4,9-hexahydropyrrolo[2,1-b]quinazolin-3a-carboxylate (23c)
A mixture of diamine 1 (102 mg, 0.835 mmol), pyruvate 19
(95 mg, 0.931 mmol), cyclopentanone (69 mg, 0.82 mmol) and
dioxane (10 mL) was heated to 100 8C for 11 days while stirring.
After evaporation of volatile components, raw oily material
(259 mg) was obtained, which was separated by CC (SiO2, CHCl3
and Et2O) to obtain two fractions: fraction 2 was pure aminal–
quinazoline 20 (78 mg, 45.3%); fraction 1 (52 mg) was separated by
preparative TLC (CHCl3/Et2O 9:1): fraction 1a (11 mg) contained
expected product 22 (ca. 6 mg, 2.8%); fraction 1b contained ca. 90%
(1H NMR) of compound 22 (4 mg, 2%).
1H NMR (300 MHz, CD3SOCD3):
d 7.41 (bs, 1H), 7.26 (bs, 1H),
7.07–6.99 (m, 2H), 6.74–6.67 (m, 2H), 4.82 (d, 1H, J = 17 Hz), 4.33
(d, 1H, J = 17 Hz), 3.66 (s, 3H), 2.95 (d, 1H, J = 14.8 Hz), 2.38 (d, 1H,
J = 14.8 Hz) ppm. 1H NMR (300 MHz, CDCl3): characteristic signals:
d
4.93 (d, 1H, J = 17 Hz), 4.42 (d, 1H, J = 17 Hz), 3.67 (s, 3H), 2.97 (d,
1H, J = 15.4 Hz), 2.42 (d, 1H, J = 14.8 Hz) ppm. 13C NMR (75 MHz,
CD3SOCD3): 170.8, 166.7, 140.9, 127.5, 126.7, 124.1 (qa,
d
J = 285.2 Hz), 118.6, 115.4, 115.1, 75 (qa, J = 29.7 Hz), 70.9, 53.2,
39.9 ppm. C14H13F3N2O4 (330.26): calcd C 50.92, H 3.90, N 8.48;
found C 50.34, H 3.97, N 8.23.
4.9.1. Methyl 2-methyl-1,2,3,4-tetrahydroquinazoline-
4.11.2. (2R*,3aS*)-Methyl-2-hydroxy-1-oxo-2-(trifluoromethyl)-
2-carboxylate (20)
1,2,3,3a,4,9-hexahydropyrrolo[2,1-b]quinazolin-3a-carboxylate (23t)
1H NMR (300 MHz, CDCl3):
d
7.0 (t, 1H, J = 7.7 Hz,), 6.86 (d, 1H,
1H NMR (300 MHz, CDCl3): characteristic signals:
d 4.81 (d, 1H,
J = 7.2 Hz), 6.67 (t, 1H, J = 7.4 Hz), 6.56 (d, 1H, J = 8.2 Hz), 4.55 (bs,
1H), 4.0 (d, 1H, J = 16.5 Hz), 3.89 (d, 1H, J = 17 Hz), 3.72 (s, 3H), 3.15
J = 17 Hz), 4.40 (d, 1H, J = 17 Hz), 3.68 (s, 3H), 2.66 (d, 1H, J = 14.8),
2.58 (d, 1H, J = 14.8) ppm.
(vbs, 1H), 1.53 (s, 3H) ppm. 13C NMR (75 MHz, CDCl3):
d 174.4p,
141.5p, 127.3n, 126n, 119.6p, 118.4n, 115n, 70p; 52.7n, 43.0p,
26.8n ppm. EIMS: m/z (rel. int.): 206 (2) [M]+, 191 (1) [MÀ15]+, 147
(100) [MÀ59]+, 131 (3), 106 (34), 104 (6), 79 (5), 78 (7), 77 (14),
42 (13).
4.12. Reaction of trifluoropyruvate 2 with 2-aminobenzamide (27a)
and cyclopentanone: attempt to prepare quinazoline 30; aldol 28
A mixture of amide 27a (438 mg, 3.217 mmol), trifluoropyr-
uvate
3.257 mmol) and dioxane (20 mL) was heated to 100 8C for
6
days while stirring (monitoring by 19F NMR). After
2
(504 mg, 3.23 mmol), cyclopentanone (274 mg,
4.9.2. 4-Hydroxy-4-methyl-2,3,3a,4,7,12-
hexahydrocyclopenta[10,20:2,3]pyrrolo[2,1-b]quinazolin-5(1H)-one
(21)
evaporation of volatile components, raw oily material (1.317 g)
was obtained that contained ca. 60% of aldol 28, but no
cyclisation product (1H NMR). Repeated purification by CC
(SiO2, 20 g, CH2Cl2) afforded aldol 28 (306 mg, 41.3%) as two
diastereoisomers (A:B = 77:23).
1H NMR (300 MHz, CDCl3): characteristic signals:
d = 7.10–7.02
(m, 2H), 6.81 (t, 1H, J = 7.7 Hz), 6.63 (d, 1H, J = 7.7 Hz), 5.07 (d, 1H,
J = 16.5 Hz), 4.31 (d, 1H, J = 16.5 Hz) ppm.
4.9.3. 4-Methylene-2,3,3a,4,6,11-
Methyl 2-hydroxy-2-(2-oxocyclopentan-1-yl)-3,3,3-trifluoropro-
panoate (28). Mp 59–60 8C (Lit. 59–60 8C [19]). Diastereoisomer
hexahydrocyclopenta[10,20:2,3]pyrrolo[2,1-b]quinazolin-5(1H)-one
(22)
A: 1H NMR (300.1 MHz, CDCl3):
d
1.64–2.37 (m, 6H, –CH2CH2CH2–),
2.95 (dd, 1H, CH, J = 8.8 and 12.1 Hz), 3.91 (s, 3H, COOCH3), 4.04
(s, 1H, OH) ppm. 19F NMR (376.6 MHz, CDCl3):
À76.81
(s, CF3) ppm. 13C NMR (75.46 MHz, CDCl3):
20.4, 24.4, 37.8 (–
1H NMR (300 MHz, CDCl3): characteristic signals:
d = 7.09–7.02
(m, 2H), 6.8 (t, 1H, J = 6.9 Hz), 6.59 (d, 1H, J = 8.2 Hz), 6.09 (d, 1H,
J = 2.2 Hz), 5.41 (d, 1H, J = 1.6), 5.14 (d, 1H, J = 17 Hz), 4.27 (d, 1H,
J = 17 Hz), 3.0 (dd, 1H, J = 2.2 and 9.3 Hz) ppm. 13C NMR (75 MHz,
d
d
CH2CH2CH2–), 51.6 (COOCH3), 54.3 (CH), 75.0 (q, C–CF3,
CDCl3):
d
166.6p, 144.1p, 141.7p, 127.5n, 126.9n, 119.6n, 117.6p,
J = 29.2 Hz); 123.2 (q, CF3, J = 287.4 Hz), 169.9 (COOCH3), 215.2
(C55O) ppm. Diastereoisomer B: 1H NMR (300.1 MHz, CDCl3):
d
117.5p, 116.3n, 80.5p, 48.5n, 38.9p, 38.1p, 34.0p, 24.5p ppm.
1.64–2.37 (m, 6H, –CH2CH2CH2–), 2.61 (dd, 1H, CH, J = 8.2 and
4.10. Cyclisation involving methyl pyruvate (19): product 20
11.5 Hz), 3.80 (s, 3H, COOCH3), 5.24 (s, 1H, OH) ppm. 19F NMR
(376.6 MHz, CDCl3):
d
À76.92 (s, CF3) ppm. 13C NMR (75.46 MHz,
A mixture of diamine 1 (101 mg, 0.827 mmol), pyruvate 19
(195 mg, 1.91 mmol) and dioxane (10 mL) was heated to 100 8C
for 11 days while stirring. After evaporation of volatile
components, raw oily material (299 mg) was obtained, which
was separated by CC (SiO2, CH2Cl2 and Et2O) to obtain quinazo-
line 20 (65 mg, 37.8%). For analytical data see Section 4.9.1.
Characteristic signals of a pyrroloquinazoline were not found in
any fraction.
CDCl3): d 23.2, 26.0, 37.9 (–CH2CH2CH2–), 50.8 (COOCH3),
53.6 (CH); 122.6 (q, CF3, J = 286 Hz), 168.1 (COOCH3), 218.7
(C55O) ppm; signal of C–CF3 was too weak.
4.13. Reaction of trifluoropyruvate (2) with 2-
(methylamino)benzamide (27b) and cyclopentanone: aldol 28
Procedure as in Section 4.9. Amide 27b (1.75 g, 11.65 mmol),
trifluoropyruvate
2 (1.818 g, 11.65 mmol), cyclopentanone
4.11. Three-component cyclisation involving trifluoropyruvate 2 and
pyruvate 19: products 20 and 23
(980 mg, 1165 mmol) and dioxane (70 mL) was heated to
100 8C for 7 days while stirring (monitoring by 19F NMR). Na
cyclisation product was detected in raw product. Aldol 28 was
obtained in yield of 723 mg (25.8%). For analytical data see
Section 4.9.
A mixture of diamine 1 (353 mg, 2,89 mmol), trifluoropyruvate
2 (456 mg, 2.922 mmol), pyruvate (19, 303 mg, 2.968 mmol) and

O. Paleta et al. / Journal of Fluorine Chemistry 157 (2014) 1–11
9
4.14. Three-component cyclisation involving 2,2-dimethoxypropane
CH2Cl2 (50 mL) and the mixture stirred until the end of reaction
(monitoring by 19F NMR). After removing volatile components
(rotary evaporator), raw product was purified by CC and then
crystallized.
(29): product 30
A mixture of 2-aminobenzamide (27a, 491 mg, 3.606 mmol),
trifluoropyruvate 2 (385 mg, 3.697 mmol), acetal 29 (385 mg,
3.697 mmol) and dioxane (15 mL) was heated to 75 8C for
5 days while stirring. After evaporation of volatile components,
raw oily material (1.158 g) was obtained that was separated
by CC (SiO2, M60, CH2Cl2/Et2O 3:2) to afford 2 fractions:
fraction 1 (151 mg) was pure stereoisomer of 30, which was
crystallized from CHCl3; fraction 2 (570 mg) was a mixture of
compounds including stereoisomers of 30 with overall yield of
25–30%.
4.16.3. Reaction of enamine 32 with hemiaminal 36:
products 38 and 41
Hemiaminal 36 (1.56 g, 5.91 mmol), enamine 32 (0.89 g,
5.91 mmol) and CH2Cl2 (100 mL). Raw product 1.08 g (imine 38:
aldol 41 = 1: 1) was separated by CC (SiO2, 60 g, CH2Cl2): fraction 2
(0.39 g) was pure aldol 41; fraction 1 (1.08 g; imine 38: aldol
41 = 1:1) was repeatedly separated by CC (SiO2, 40 g, CH2Cl2) to
afford aldol 41 (0.68 g) and unstable imine 38, which was
converted to aldol 41 to afford total yield 1.07 g (75.4%) as a
mixture of two diastereoisomers (A/B 74/26); mp 59–60 8C (also
Ref. [23]).
Isolated stereoisomer of 2-hydroxy-2-(trifluoromethyl)-
2a,3,4,5-tetrahydrocyclopenta[10,20:2,3]-pyrrolo[1,2-a]quinazo-
lin-1,7(2H,6H)-dione (30). IR: 3366 (m), 3200–2800 (broad), 1705
(s), 1679 (s), 1604 (m) cmÀ1. 1H NMR (300 MHz, CD3SOCD3):
d
9.09
Methyl 2-[2-(benzylimino)cyclopentan-1-yl]-2-hydroxy-3,3,3-tri-
fluoropropanoate (38). Diastereoisomer A: 1H NMR (300 MHz,
CDCl3): d 1.59–2.52 (m, 6H, –CH2CH2CH2–), 3.28 (dd, 1H, CH, J = 6.7
(s, 1H), 8.02 (d, 1H, J = 7.7 Hz), 7.95 (d, 1H, J = 7.7 Hz), 7.79 (s, 1H),
7.7 (t, 1H, J = 7.4 Hz), 7.4 (t, 1H, J = 7.1 Hz), 2.62 (d, 1H, J = 13.7 Hz),
ꢀ2.5 (m, 1H, overlap with DMSO), 1.57 (s, 3H) ppm. 1H NMR
and 8.1 Hz), 3.72 (s, 3H, COOCH3), 4.42 (d, 1H, CH2, J = 15.1 Hz),
4.51 (d, 1H, CH2, J = 15.1 Hz), 4.36 (s, 1H, OH), 7.18–7.89 (m, 5H, Ph)
(300 MHz, CDCl3): characteristic signals: d 2.63 (d, 1H, J = 14.3 Hz),
2.75 (d, 1H, J = 14.3 Hz) ppm. 13C NMR (75 MHz, CD3SOCD3):
164.8p, 160.8p, 134.4p, 133.6n, 128n, 126n, 123.8p (qa,
J = 284.6 Hz), 120.3p, 120.2n, 75.6p (qa, J = 30.3 Hz), 71.5p,
ppm. 19F NMR (376.6 MHz, CDCl3):
(75.5 MHz, CDCl3):
d
À75.49 (s, CF3) ppm. 13C NMR
d
22.1, 25.3, 28.9 (–CH2CH2CH2–), 48.5
(COOCH3), 52.1 (CH), 56.6 (CH2), 79.4 (qa, C–CF3, J = 29 Hz),
123.5 (qa, CF3, J = 287 Hz), 126.1, 127.1, 127.9, 139.6 (Ph), 168
(COOCH3) ppm.
41.7p, 28.1n ppm. 19F NMR (376 MHz, CD3SOCD3):
ppm. C13H11F3N2O3 + CHCl3 (419.614): calcd C 40.07, H 2.88, N
6.98; found C 40.29, H 2.99, N 7.22.
d
À78.2 (s)
Diastereoisomer B: 1H NMR (300 MHz, CDCl3):
d 1.59–2.52 (m,
6H, –CH2CH2CH2–), 3.69 (s, 3H, COOCH3), 7.18–7.89 (m, 5H, Ph)
ppm; the signals of CH, CH2 and OH were not assigned due to their
4.15. Cyclisation involving 2,2-dimethoxypropane (29): product 31
low intensity. 19F NMR (376.6 MHz, CDCl3):
d
À75.51 (s, CF3) ppm.
A mixture of diamine 1 (436 mg, 3.569 mmol), trifluoropyr-
uvate 2 (549 mg, 3.518 mmol), acetal 29 (374 mg, 3.591 mmol)
and dioxane (15 mL) was heated to 100 8C for 5 days while
stirring. After evaporation of volatile components, raw oily
material (1.058 g) was obtained, which contained cis-31 and
trans-31 in a ratio 60:40. The material was first purified using CC
(SiO2, CHCl3 and Et2O) and then that was separated by CC (SiO2,
CH2Cl2/acetone 9:1) to afford 294 mg (29.2%) of a mixture of
isomers 31c and 31t in ratio 59:41. For analytical data see Ref.
[9].
13C NMR (75.5 MHz, CDCl3): the signals were not assigned due to
their low intensity.
Methyl 2-hydroxy-2-(2-oxocyclopentan-1-yl)-3,3,3-trifluoropro-
panoate (41). Diastereoisomer A: 1H NMR (300 Hz, CDCl3):
d
1.64–2.37 (m, 6H, –CH2CH2CH2–), 2.95 (dd, 1H, CH, J = 8.8 and
12.1 Hz), 3.91 (s, 3H, COOCH3), 4.04 (s, 1H, OH) ppm. 19F NMR
(376.6 MHz, CDCl3):
CDCl3):
d
À76.81 (s, CF3) ppm. 13C NMR (75.5 MHz,
d
20.4, 24.4, 37.8 (–CH2CH2CH2–), 51.6 (COOCH3), 54.3
(CH), 75 (qa, C–CF3, 2JCF 29.2 Hz), 123.1 (qa, CF3, J = 287 Hz), 169.9
(COOCH3), 215.2 (C55O) ppm.
Diastereoisomer B: 1H NMR (300 Hz, CDCl3):
d 1.64–2.37 (m,
4.16. Reactions of enamines 32–35 with hemiaminal 36
4.16.1. Preparation of benzylamine hemiaminal 36
6H, –CH2CH2CH2–), 2.61 (dd, 1H, CH, J = 8.2 and 11.5 Hz), 3.80 (s,
3H, COOCH3), 5.24 (s, 1H, OH) ppm. 19F NMR (376.6 MHz, CDCl3):
d
= À76.92 (s, CF3) ppm. 13C NMR (75.5 MHz, CDCl3):
d 23.2, 26,
Under argon,
a
solution of trifluoropyruvate
2
(0.95 g,
37.9 (–CH2CH2CH2–), 50.8 (COOCH3); 53.6 (CH), 122.6 (qa, CF3,
J = 285.7 Hz), 168.1 (COOCH3), 218.7 (C55O) ppm. The signal for C–
CF3 was not assigned due to low intensity.
5.8 mmol) in CH2Cl2 (5 mL) was dropwise added at rt while
stirring to benzylamine (0.62 g, 5.8 mmol) in CH2Cl2 (50 mL)
and the mixture stirred until the end of reaction (monitoring by
19F NMR). After removing volatile components (rotary evapora-
tor), pure heminal 36 was obtained (1.29 g, 80.3%), mp 54–
56 8C.
4.16.4. Reaction of enamine 33 with hemiaminal 36: product 37
Hemiaminal 36 (1.52 g, 5.79 mmol), enamine 33 (0.96 g,
5.79 mmol) and CH2Cl2 (100 mL). Raw product was purified by
CC (SiO2, 60 g, CHCl3/Et2O 9:1) and then crystallized (hexane/
CH2Cl2 4:1) to obtain pure indolone 37 (1.35 g, 74.8%), mp 134–
138 8C.
Methyl 2-benzylamino-3,3,3-trifluoro-2-hydroxypropanoate (36).
1H NMR (300.07 MHz, CDCl3):
d 2.46 (bs, 1H, OH nebo NH), 3.71 (d,
1H, CH2, J = 12.9 Hz), 3.81 (s, 3H, COOCH3), 3.95 (d, 1H, CH2,
J = 12.9 Hz), 4.52 (bs, 1H, OH or NH), 7.23–7.38 (m, 5H, Ph-H) ppm.
1-Benzyl-3-hydroxy-3-(trifluoromethyl)-1,3,4,5,6,7-hexahydroin-
dol-2-one (37). 1H NMR (300 MHz, CDCl3):
d 1.48–1.69 (m, 2H,
19F NMR (376.6 MHz, CDCl3):
(75.46 MHz, CDCl3):
d
À80.5 (s, CF3) ppm. 13C NMR
d
45.91 (CH2), 54.32 (COOCH3), 85.43 (q, C–
CH2), 1.72–1.86 (m, 2H, CH2), 2.05–2.19 (m, 3H, CH2), 2.22–2.29
2
CF3, J = 31.2 Hz), 122.19 (q, CF3, J = 287.6 Hz); 127.47, 128.31,
128.46, 138.45 (Ph); 168.93 (COOCH3) ppm. EIMS: m/z (rel. int.):
204 (2) [MÀ59]+, 195 (1.2), 129 (2.6), 115 (4.4), 107 (65.7); 106
(87.6), 99 (7.6); 97 (7.3), 91 (35), 79 (33), 78 (18), 77 (19.5), 69
(100), 65 (7), 60 (5.3), 59 (92), 53 (6), 52 (5.6), 51 (17), 50 (17).
(m, 1H, CH2), 4.53 (d, 1H, CH2, J = 16 Hz), 4.74 (d, 1H, CH2, JHH
16 Hz), 4.94 (bs, 1H, OH), 7.14–7.36 (m, 5H, Ph-H) ppm. 19F NMR
(376.6 MHz, CDCl3):
CDCl3):
d
À79.84 (s, CF3) ppm. 13C NMR (75.5 MHz,
d
19.8 (CH2), 21.6 (CH2), 21.7 (CH2), 21.9 (CH2), 43.3 (CH2),
19.8 (CH2), 76.8 (qa, C–CF3, J = 31 Hz), 111.8 (C55), 123.2 (qa, CF3,
J = 286 Hz), 126.5, 127.4, 128.6, 143.7 (Ph), 136.3 (C55), 173.6
(C55O) ppm. C16H16F3NO2 (311.11) calcd: C 61.73, H 5.18, N 4.50;
4.16.2. General procedure
Under argon, a solution of enamine 32–35 in CH2Cl2 (50 mL)
was dropwise added at r.t. while stirring to hemiaminal 36 in
found:
Supplementary data, p. 8.
C 61.46, H 5.24, N 4.66. For X-ray structure see

10
O. Paleta et al. / Journal of Fluorine Chemistry 157 (2014) 1–11
4.16.5. Reaction of enamine 34 with hemiaminal 36:
products 39 and 42
continued at rt for 3 days (check by 19F NMR: no 44, one product).
The mixture was then stirred at 50 8C for 18 h (check by 19F NMR:
no change, no 44). After removing volatile components, the residue
was purified by CC (SiO2, CH2Cl2) to obtain a mixture of 3
components (760 mg), in which were identified (NMR and GC–MS)
aniline, benzylamine and aldol 43 (see Section 4.16.6.)
Hemiaminal 36 (1.69 g, 6.41 mmol), enamine 34 (1.2 g,
6.41 mmol) and CH2Cl2 (100 mL). Raw product (1.08 g, imine
39: aldol 42 = 15:1) was separated by CC (SiO2, 60 g, CH2Cl2-
hexane 9:1): fraction 2 (0.19 g) was pure aldol 42; fraction 1
(1.06 g; imine 39: aldol 42 = 7:1) was repeatingly separated by CC
(SiO2, 40 g, CH2Cl2-hexane 9:1) to afford aldol 43 (0.42 g) and
mixture of imine 39: aldol 42 = 1:4 (0.45 g). After conversion of
unstable imine 39 to aldol 42, the second was obtained in total
yield of 0.91 g (51.4%), mp 85–87 8C (also Ref. [23]).
Acknowledgements
The research was supported by the Grant Agency of the Czech
Republic (Project No. 203/02/03), the Czech Ministry of Education
(Projects MSM 6046137301) and by the Institute of Chemical
Technology, Prague.
Methyl 4-(benzylimino)-2-hydroxy-4-phenyl-2-(trifluoromethyl)-
butanoate (39). 1H NMR (300 MHz, CDCl3):
d 3.08 (d, 1H, CH2,
J = 17.3 Hz), 3.49 (d, 1H, CH2, J = 17.6 Hz), 4.44 (d, 1H, CH2,
J = 14.6 Hz), 4.51 (d, 1H, CH2, J = 14.6 Hz), 3.63 (s, 3H, COOCH3),
6.76 (s, 1H, OH), 7.14–7.96 (m, 10H, Ph) ppm. 19F NMR (376.6 MHz,
Appendix A. Supplementary data
CDCl3):
d
À79.4 (s, CF3) ppm. 13C NMR (75.5 MHz, CDCl3):
d 41
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.jfluchem.2013.
10.017.
(CH2), 53.1 (COOCH3), 56.3 (CH2), 77.7 (qa, C–CF3, J = 29.1 Hz), 123
(qa, CF3, J = 285 Hz), 126.1, 126.9, 127.5, 128.4, 128.8, 136, 138.8
(2xPh), 169.9 (COOCH3) ppm.
Methyl 2-hydroxy-4-oxo-4-phenyl-2-(trifluoromethyl)butanoate
References
(42). 1H NMR (300 MHz, CDCl3):
d
3.70 (d, 1H, CH2, J = 17.6 Hz),
3.77 (d, 1H, CH2, J = 17.6 Hz), 3.91 (s, 3H, COOCH3), 4.25 (s, 1H, OH),
7.47–7.95 (m, 5H, Ph) ppm. 19F NMR (376.6 MHz, CDCl3):
À79.5
(s, CF3) ppm. 13C NMR (75.5 MHz, CDCl3):
40.7 (CH2), 54.2
´
ˇ
´ˇ
[1] (a) O. Paleta, B. Dolensky, J. Palecek, J. Kvıcala, Eur. J. Org. Chem. (2013) 1262–
d
1270;
ˇ
ˇ
´ ˇ
(b) B. Dolensky´, J. Kvı´cala, O. Paleta, J. Cejka, J. Ondracek, Tetrahedron Lett. 37
(1996) 6939–6942.
d
(COOCH3), 75 (qa, C–CF3, J = 29.2 Hz), 123.2 (q, CF3, J = 286 Hz),
128.2, 128.8, 134.1, 135.8 (Ph), 169.4 (COOCH3), 194.8 (C55O) ppm.
[2] (a) A. Do¨mling, W. Ho¨rl, Endeavour 18 (1994) 115–122;
(b) L.F. Tietze, Chem. Rev. 96 (1996) 115–136;
(c) A. Do¨mling, I. Ugi, Angew. Chem. Int. Ed. Engl. 39 (2000) 3168–3210;
´
(d) D.J. Ramon, M. Yus, Angew. Chem. Int. Ed. Engl. 44 (2005) 1602–1634;
4.16.6. Reaction of enamine 35 with hemiaminal 36:
products 40 and 43
(e) J. Zhu, H. Bienayme´ (Eds.), Multicomponent Reactions, Wiley-VCH, Weinheim,
2005;
(f) A. Dondoni, A. Massi, Acc. Chem. Res. 39 (2006) 451–463;
(g) G. Guillena, D.J. Ramo´n, M. Yus, Tetrahedron Asym. 18 (2007) 693–700;
(h) C.J. Chapman, C.G. Frost, Synthesis (2007) 1–21;
(j) B.B. Toure´, D.G. Hall, Chem. Rev. 109 (2009) 4439–4486;
(k) R.V.A. Orru, E. Ruijter (Eds.), Synthesis of Heterocycles via Multicomponent
Reactions, Vols. 1 and 2, Springer, Heidelberg, 2010;
(m) D. Bonne, T. Constantineux, Y. Coquerel, J. Rodriguez, Chem. Eur. J. 19 (2013)
2218–2231.
Hemiaminal 36 (2.07 g, 7.88 mmol), enamine 35 (2.07 g,
7.88 mmol) and CH2Cl2 (100 mL). Raw product (imine 40: aldol
43 = 7:1) was separated by CC (SiO2, 60 g, CH2Cl2-hexane 9:1):
fraction 2 (0.41 g) was pure aldol 43; fraction 1 (1.42 g; imine 40:
aldol 43 = 7:1) was repeatedly separated by CC (SiO2, 40 g, CH2Cl2-
hexane 9:1) to afford aldol 43 (0.39 g) and mixture of imine 40:
aldol 43 = 1:4 (0.45 g). After conversion of unstable imine 40 to
aldol 43, the second was obtained in total yield of 1.19 g (71.9%),
mp 61–63 8C (also Ref. [23]).
[3] Hantzsch synthesis of dihydropyridines:
(a) A. Hantzsch, Chem. Ber. 14 (1881) 1637–1638;
(b) U. Eisner, J. Kuthan, Chem. Rev. 72 (1972) 1–42;
(c) J. Kuthan, A. Kurfu¨rst, Ind. Eng. Chem. Prod. Res. Rev. 21 (1982) 191–261;
Methyl 4-(benzylimino)-2-hydroxy-2-(trifluoromethyl)pentano-
ˇ
ˇ
ˇ
(d) M. Filipan-Litvic, M. Litvic, I. Cepanec, V. Vinkovic, Molecules 12 (2007) 2546–
2558.
ate (40). 1H NMR (300.07 MHz, CDCl3):
d 2.11 (s, 3H, CH3), 2.74
[4] Biginelli synthesis of dihydropyrimidines:
(d, 1H, CH2, J = 17.9 Hz), 3.11 (d, 1H, CH2, J = 17.9 Hz), 3.72 (s, 3H,
COOCH3), 3.12 (d, 1H, CH2, J = 14.4 Hz), 3.69 (d, 1H, CH2,
J = 14.4 Hz), 7.06–7.62 (m, 5H, Ph) ppm; group OH was assigned.
19F NMR (376.6 MHz, CDCl3): À79.21 (s, CF3) ppm. 13C NMR
(a) P. Biginelli, Gazz. Chim. Ital. 23 (1893) 360–369;
(b) C.O. Kappe, Tetrahedron 49 (1993) 6937–6963;
(c) C.O. Kappe, Acc. Chem. Res. 33 (2000) 879–888.
[5] Mannich synthesis of aminoketones:
(a) M. Tramontini, Synthesis (1973) 703–707;
(b) M. Arend, B. Westermann, N. Risch, Angew. Chem. Int. Ed. Engl. 37 (1998)
1044–1070;
(75.46 MHz, CDCl3):
d 28.26 (CH3), 45.98 (CH2), 48.34 (CH2), 51.67
(COOCH3), 77.15 (q, C–CF3, J = 29.9 Hz), 121.18 (q, CF3, J = 285 Hz);
126.6, 127.2, 128.2, 138.5 (all Ph), 167.7 (COOCH3) ppm.
Methyl 2-hydroxy-4-oxo-2-(trifluoromethyl)pentanoate (43). 1H
(c) A. Co´rdoba, Acc. Chem. Res. 37 (2004) 102–112;
(d) M.M.B. Marques, Angew. Chem. Int. Ed. Engl. 45 (2006) 348–352.
[6] (a) P.S.V. Prasanna, P. Gunasekaran, S. Perumal, Tetrahedron Lett. 54 (2013)
3740–3743;
NMR (300.07 MHz, CDCl3):
d 2.17 (s, 3H, CH3), 3.11 (d, 1H, CH2,
(b) N. Yohreh, A. Aliyadeh, ACS Combinat. Sci. 15 (2013) 278–286;
(c) J. Li, M.-S. Yan Yu, B. Tu, S.L. Jiang, S.-J. Wang, Tu, Org. Biomol. Chem. 10 (2012)
5361–5365;
(d) J. Zhang, M. Zhang, W. Cao, L. Song, Q. Qian, J. Tan, M.J. Shao, Fluorine Chem.
130 (2009) 488–492.
J = 17.6 Hz), 3.21 (d, 1H, CH2, J = 17.6 Hz), 3.88 (s, 3H, COOCH3),
4.11 (s, 1H, OH) ppm. 19F NMR (376.6 MHz, CDCl3): À79.71 (s, CF3)
ppm. 13C NMR (75.46 MHz, CDCl3):
d 30.3 (CH3), 44.8 (CH2), 54.0
(COOCH3), 78.9 (q, C–CF3, J = 30.3 Hz), 122.9 (q, CF3, J = 286 Hz),
169.1 (COOCH3), 203.5 (C55O) ppm. EIMS: m/z (rel. int.): 215 (0.3)
[M+1]+, 214 (0.3) [M]+, 181(2), 155 (19) [MÀ59]+, 139 (8), 101 (3)
93 (33), 69 (10), 66 (16), 59 (10), 43 (100).
[7] (a) A. Aliyadeh, A. Reyvanian, Helv. Chim. Acta 95 (2012) 858–862;
(b) J. Sun, Y. Sun, H. Gao, C.-G. Yan, Eur. J. Org. Chem. (2011) 6952–6956;
(c) P. Amutha, S. Nagarajan, Helv. Chim. Acta 93 (2010) 430–434;
(d) J.F. Liu, P. Ye, K. Sprague, K. Sargent, D. Yohannes, C.M. Baldino, C.J. Wilson, S.C.
Nag, Org. Lett. 7 (2005) 3363–3366;
(e) S.J. Tu, B. Jiang, C.S. Yao, H. Jiang, J.Y. Zhang, R.H. Jia, Y. Zhang, Synthesis (2007)
1366–1372;
4.17. Attempt at four-component domino cyclisation involving
(f) R.S. Kumar, S.M. Rajesh, S. Perunal, D. Banerjee, P. Yogeeswari, D. Sriram, Eur. J.
Med. Chem. 45 (2010) 411–422.
trifluoropyruvate (2) to give 44
[8] (a) R. Heydari, B. Tahmipour, N.A. Torbati, C. Gaiff, M. Yizaadini, Synth. Commun.
43 (2013) 2031–2041;
Apparatus as in Section 4.4. To a flask (25 mL), solutions of
reactants in dioxane (5 mL) were dropwise added at rt while
stirring in the order as follows: trifluoropyruvate (2, 539 mg,
3.45 mmol), benzylamine (374 mg, 3.49 mmol), aniline (322 mg,
3.46 mmol) and acetone (202 mg, 3.48 mmol) and stirring was
(b) R.W. Armstrong, A.P. Combs, P.A. Tempest, S.D. Brown, T.A. Keating, Acc.
Chem. Res. 29 (1996) 123–131;
(c) S.L. Daw, J.J. McNally, M.A. Youngman, Curr. Med. Chem. 6 (1999) 255–
270;
(d) L. Weber, Combinat. Chem. 7 (2002) 143–147.

O. Paleta et al. / Journal of Fluorine Chemistry 157 (2014) 1–11
11
ˇ
ˇ
ˇ´
´
[9] B. Dolensky´, J. Kvı´cala, O. Paleta, J. Lang, H. Dvorakova, J. Cejka, Magn. Res. Chem.
(b) F. Johnson, G. Duquette, A. Whitehead, L.C. Dorman, Tetrahedron 30 (1974)
3241–3249;
48 (2010) 375–385.
ˇ
[10] B. Dolensky´, J. Kvı´cala, O. Paleta, J. Fluorine Chem. 126 (2005) 745–751.
(c) H. Mazarguil, A. Lattes, Tetrahedron Lett. (1971) 975–984.
ˇ
ˇ
[11] (a) F. Csende, G. Bernath, Z. Boecskei, P. Sohar, G. Stajer, Heterocycles 45 (1997)
323–329;
[16] B. Dolensky´, J. Kvı´cala, J. Palecek, O. Paleta, J. Fluorine Chem. 115 (2002)
67–74.
(b) M. Yamato, Y. Takeuchi, Chem. Pharm. Bull. 30 (1982) 1036–1040;
(c) S.D. Sharma, V. Kaur, P. Sharma, Ind. J. Chem. Sect. B. 32 (1993) 517–524.
[12] (a) I.L. Knunyants, V.V. Shokhina, V.V. Tyuleneva, Dokl. Akad. Nauk SSSR 169
(1966) 594–601;
[17] (a) H.Z. Kolbe, Prakt. Chem 30 (1884) 467–482;
(b) L.A. Errede, P.D. Matinucci, J.J. McBrady, J. Org. Chem. 45 (1980) 3009–3017.
[18] W.E. Coyne, J.W.J. Cusic, Med. Chem. 11 (1968) 1208–1213.
[19] T. Ishihara, H. Shinjo, Y. Inoue, T.J. Ando, Fluorine. Chem. 22 (1983) 1–20.
[20] (a) R.F. Paercell, F.P. Hauck Jr., J. Org. Chem. 28 (1963) 3468–3473;
(b) V.I. Saloutin, I.A. Piterskikh, K.I. Pashkevich, Izv. Akad. Nauk SSSR. Ser. Khim.
(1986) 625–632.
ˇ
´
ˇ
(b) M. Procha´zka, V. Krest’anova, M. Palecek, K. Pecka, Collect. Czech. Chem.
[13] (a) K. Funabiki, M. Nojiri, M. Matsui, K. Shibata, Chem. Commun. (1998) 2051–
2052;
Commun. 25 (1970) 3813–3817.
[21] (a) Z. Eckstein, E. Grochowski, T. Urbanski, Bull. Acad. Pol. Sci. Ser. Sci. Chim. 7
(1959) 289–302;
(b) J.S.L. Ibaceta-Lizana, R. Iyer, A.H. Jackson, P.V.R. Shannon, J. Chem. Soc. Perkin
Trans. 2 (1978) 733–742.
[22] (a) M.E. Munk, Y.K. Kim, J. Am. Chem. Soc. 86 (1964) 2213–2217;
(b) Y.G. Safina, G.S. Malkova, R.A. Cherkasov, Zh. Obshch. Khim. 61 (1991) 620–
634;
(b) K. Funabiki, A. Isomura, Y. Yamaguchi, W. Hashimoto, K. Matsunaga, K.
Shibata, M.J. Matsui, Chem. Soc. Perkin Trans. 1 (2001) 2578–2582;
(c) K. Funabiki, K. Matsunaga, M. Nojiri, W. Hashimoto, H. Yamamoto, K. Shibata,
M. Matsui, J. Org. Chem. 68 (2003) 2853–2860;
ˇ
(d) J. Palecek, O. Paleta, Synthesis (2004) 521–524.
[14] (a) G. Illuminati, L. Mandolini, Acc. Chem. Res. 14 (1981) 95–102;
(b) J.E. Baldwin, J. Chem. Soc. Chem. Commun. (1976) 738–741.
[15] (a) O. Exner, in: N.B. Chapman, J. Shorter (Eds.), Correlation Analysis in Chemis-
try–Recent Advances, Plenum Press, New York, 1978, p. 439;
Chem. Abstr. 115 (1991) 136225.
[23] H. Ahlbrecht, W. Raab, Synthesis (1980) 320–322.
ˇ
´
[24] O. Paleta, J. Palecek, B. Dolensky, J.Fluorine Chem. 111 (2001) 175–184.