Acridine synthesis by the reaction of 2-(perfluoroalkyl)aniline with ortho-alkyl-substituted aromatic grignard reagent: Mechanistic studies

Article abstract of DOI:10.1071/CH14401

The reaction of 2-(perfluoroethyl)aniline and its higher perfluoroalkyl analogues with an arylmagnesium bromide substituted with a methyl or ethyl group at the ortho position furnishes an acridine containing a shorter perfluoroalkyl group at the 9-position and devoid of the methyl or ethyl group of the Grignard reagent. Yields are in the range of 46-93%. The intermediary of a substituted aza-ortho-xylylene has been postulated for related transformations in the literature previously, but this intermediate product has never been isolated. As part of this work, the labile product E-27 (half-life of 6h at 23°C) was isolated for the first time and characterized by infrared spectroscopy, electron impact mass spectrometry, fast atom bombardment mass spectrometry, and 1H NMR. Experimental evidence was also obtained regarding the elimination of the ortho-alkyl group of the Grignard reagent during the course of the reaction as an alcohol.

Full text of DOI:10.1071/CH14401

RESEARCH FRONT
CSIRO PUBLISHING
Aust. J. Chem. 2015, 68, 196–202
Full Paper
http://dx.doi.org/10.1071/CH14401
Acridine Synthesis by the Reaction of 2-(Perfluoroalkyl)
aniline with ortho-Alkyl-Substituted Aromatic Grignard
Reagent: Mechanistic Studies
A E
,
B
C
Lucjan Strekowski, Jianguo Zhang, Jarosław Sa˛czewski,
and Ewa Wolin´ska
D
A
Department of Chemistry, Georgia State University, Georgia 30302-4098, USA.
Department of Biochemistry and Molecular Biology, The University of Texas Medical
Branch, 301 University Blvd., Galveston, Texas 77555-0635, USA.
B
C
Department of Chemical Technology of Drugs, Medical University of Gdan´sk,
Al. Gen. J. Hallera 107, 80-416 Gdan´sk, Poland.
D
Department of Chemistry, Siedlce University, 08-110 Siedlce, Poland.
E
Corresponding author. Email: Lucjan@gsu.edu
The reaction of 2-(perfluoroethyl)aniline and its higher perfluoroalkyl analogues with an arylmagnesium bromide
substituted with a methyl or ethyl group at the ortho position furnishes an acridine containing a shorter perfluoroalkyl
group at the 9-position and devoid of the methyl or ethyl group of the Grignard reagent. Yields are in the range of 46–93 %.
The intermediary of a substituted aza-ortho-xylylene has been postulated for related transformations in the literature
previously, but this intermediate product has never been isolated. As part of this work, the labile product E-27 (half-life of
6 h at 238C) was isolated for the first time and characterized by infrared spectroscopy, electron impact mass spectrometry,
1
fast atom bombardment mass spectrometry, and H NMR. Experimental evidence was also obtained regarding the
elimination of the ortho-alkyl group of the Grignard reagent during the course of the reaction as an alcohol.
Manuscript received: 20 June 2014.
Manuscript accepted: 9 July 2014.
Published online: 22 September 2014.
Introduction
Novel chemistry has been observed for aryl Grignard
reagents reacting as nucleophilic bases with 2-(perfuoroalkyl)
anilines. Thus, the treatment of 2-(trifluoromethyl)aniline (1)
with phenylmagnesium bromide furnishes a tetraarylmethane
3 as the major product[12] (Scheme 1). On the other hand,
the reaction of one molecule of 1 with two molecules of
2-ethylphenylmagnesium bromide (4) yields a substituted
acridine 5 that is devoid of the ethyl group from one of the
molecules of the Grignard reagent that reacted.[13] In a similar
way, the reaction of 1 with 2,6-dimethylphenylmagnesium
bromide gives an acridine 7, the molecular formula of which is
shorter by one carbon (methyl group) than that of the molecular
combination of 1 and two molecules of the Grignard reagent
required to form product 7. This report relates to the reactions of
higher 2-(CnF2nþ1)anilines with ortho-alkyl-substituted aroma-
tic Grignard reagents that produce 9-(Cn–1F2n–1)acridines that do
not contain the alkyl group of the starting Grignard reagent.[14]
The previously published chemistry of 1 is included for compar-
ison to understand better the reactivity pattern of structurally
different substrates. The general synthesis of 9-(perfluoroalkyl)
acridines is illustrated in Schemes 2 and 3 by our previously
unpublished results. The acridines reported recently have been
adequately characterized[14] and are not included in this work.
The mechanism elucidated as part of this research is shown in
Schemes 4 and 5.
2-(Trifluoromethyl)aniline is a commercial product, and its
analogues containing various substituents at the ring can easily
be obtained by the Ullmann condensation of a substituted
2-bromoaniline or 2- iodoaniline with trifluoromethyl iodide
in the presence of copper bronze,[1,2] or by reductive tri-
fluoromethylation of an aniline with trifluoromethyl iodide in
the presence of zinc and sulfur dioxide.[3,4] The latter reaction
is ortho regioselective.[4] Other perfluoroalkyl derivatives can
conveniently be prepared similarly using a commercial
perfluoroalkyl iodide. The accessibility of a variety of
2-(perfluoroalkyl)anilines has stimulated their application in the
synthesis of heterocyclic compounds in the process called
anionic activation of the trifluoromethyl group and the benzylic
position of higher perfluoroalkyl substituents. More specifical-
ly, the ionization of the amino group followed by elimination of
fluoride result in the generation of a non-aromatic aza-ortho-
xylylene intermediate product that can undergo aromatization
by addition reaction with a nucleophile (see below). Numerous
practical methods for the synthesis of heterocyclic compounds
have been developed based on the anionic activation of the
trifluoromethyl group in the presence of various nucleophilic
bases. This chemistry has been extensively reviewed by us[5–8]
and others.[9–11] In contrast, the anionic activation of
2-(perfluoroalkyl)anilines has received much less attention.
Journal compilation Ó CSIRO 2015
www.publish.csiro.au/journals/ajc

Acridine Synthesis
197
MgBr
nature of the aryl groups, a single molecule may exist in a large
number of conformational isomers. The two methylene moieties
in 11 are anisochronous at 238C in the 1H NMR time scale, and
they give rise to two clearly defined quartets, indicating inde-
pendent coupling with the adjacent methyl groups. The observed
Ph
Ph
Ph
2
[ref. 12]
NH2
1
absorption pattern in the H NMR spectrum of 11 recorded at
3
(25 %)
238C or lower gives rise to a broad absorption as the temperature
increases, eventually becoming a single sharp quartet at the
coalescence temperature of 558C. In a similar way, the two
anisochronous methyl groups in molecular propeller 12 show
two singlets at 238C that coalesce to one singlet at 508C.
It should be noted that the structure of acridine 10 lacks the
ortho-alkyl group of the starting Grignard reagents 4 and 9.
The same reactivity pattern can be seen in the synthesis of
9-(trifluoromethyl)acridines 14 and 15 obtained by treatment of
aniline 8 with the respective reagents 6 and 13. More specifi-
cally, the products 14 and 15 are devoid of one methyl group of
the starting Grignard reagent.
MgBr
Et
Et
CF3
4
[ref. 13]
NH2
N
1
5
(66 %)
MgBr
Me
Me
Me
Me
We have shown previously that the treatment of
2-(perfluoroethyl)aniline (8), 2-(perfluoropropyl)aniline (16),
or 2-(perfluorobutyl)aniline (19) with unsubstituted phenylmag-
nesium bromide produces cleanly the corresponding molecular
propellers.[15] As shown in Scheme 2, a mixture of a molecular
propeller and an acridine is formed by the reaction of
2-(perfluoroethyl)aniline (8) with 2-alkylphenylmagnesium
bromide, but the acridine product is formed exclusively upon
treatment of 8 with 2,6-dimethylphenylmagnesium reagents.
This structural pattern of the formation of the acridine
system exclusively persists for the reactions of higher
2-(perfluoroalkyl)anilines including perfluoropropyl 16, per-
fluorobutyl 19, and perfluorohexyl 20 derivatives (Scheme 3)
with 2-methylphenylmagnesium and 2,6-dimethylphenyl-
magnesium reagents.
Me
6
[ref. 13]
N
7
(35 %)
Scheme 1.
CF3
N
R
CF3
NH2
R
ꢁ
10 (46 % from 4)
(47 % from 9)
11: R ꢀ Et (46 % from 4)
12: R ꢀ Me (44 % from 9)
Mechanistic Studies
This work was highly frustrating to us for several years. On the
one hand, it was highly rewarding to successfully elucidate the
practical scope and limitations of this new efficient synthesis of
acridines, so that a new acridine structure could be designed and
synthesized. On the other hand, for several years we had no clue
about the mechanism. The mechanism elucidated as part of
this work is given in Scheme 4. A particular reaction between
2-(perfluorobutyl)aniline (19) and 2,6-dimethylphenylmagne-
sium bromide (6) is exemplified because this reaction furnishes
acridine 33 in high yield i.e. .92 %.
MgBr
R
MgBr
4: R ꢀ Et
9: R ꢀ Me
Me
Me
CF3 Me
6: R ꢀ H
C2F5
13: R ꢀ Me
R
N
R
NH2
Previous studies conducted by us and other scientists on
analogous reaction systems clearly suggested that the first step is
deprotonation of the aniline, which generates an anionic inter-
mediate product, such as 24, shown in Scheme 4. Elimination of
fluoride from the benzylic position of 24 is then followed by
nucleophilic addition of the Grignard substrate to the resultant
ortho-quinoid intermediate 25. This reaction gives rise to the
aromatic adduct 26 that can undergo another elimination of
fluoride in a way similar to the initial reaction mentioned above
to give two cis and trans isomers Z-27 and E-27. It is suggested
that these isomers can undergo equilibration under basic condi-
tions by deprotonation of a methyl group and the intermediary of
29. It is the intermediate product E-28 (magnesium derivative
of E-27) that undergoes electrocyclization to generate the non-
aromatic acridine derivative 30. The methyl shift in 30 to
generate N-methyl derivative 31 concludes the reaction
sequence before workup. Quenching the mixture with water
followed by oxidation of the resultant dihydroacridine with
8
14: R ꢀ H (74 %)
15: R ꢀ Me (80 %)
Scheme 2.
Results and Discussion
Synthesis
The reactions of alkyl-substituted phenylmagnesium reagents
with 2-(perfluoroethyl)aniline (8) are shown in Scheme 2.
Treatment of 8 with 2-ethylphenylmagnesium bromide (4)
furnished an equimolar mixture of acridine 10 and a molecular
propeller 11. The same acridine 10 and an analogous mole-
cular propeller 12 were obtained by the reaction of 8 with
2-tolylmagnesium bromide (9). Molecular propellers exist in
chiral helical conformations in which the helicity and the cor-
related rotation of the planar substituents (blades) are imposed
by severe steric hindrance in the molecule.[15] Depending on the

198
L. Strekowski et al.
MgBr
F5C2
C3F7
R2
Me
9
: R1 ꢀ R2 ꢀ H
13: R1 ꢀ R2 ꢀ Me
21: R1 ꢀ Me; R2 ꢀ H
N
N
Me
R1
17 (93 %)
22 (86 %)
16
16
ꢁ
9
19
20
ꢁ
ꢁ
21
13
CnF2nꢁ1
NH2
ꢁ
13
F5C2
Me
F11C5
Me
16: n ꢀ 3
19: n ꢀ 4
20: n ꢀ 6
N
Me
N
Me
23 (55 %)
18 (67 %)
Scheme 3.
Me
MgBr
BrMg
Me
Me
Me
C3F7
F
C4F9
NH2
C4F9
6
ꢂMgBrF
ꢂ78ꢃC
NH
NH
19
24
25
MgBr
Me
Me
Me
F
F7C3
Me
Me
F7C3
ꢂMgBrF
ꢁ
C3F7
Me
NH
NH
N
R
MgBr
E-27
26
Z-27: R ꢀ H
Z-28: R ꢀ MgBr
ꢂMgBrF
Me
F7C3
Me
Me
CH2
C3F7
C3F7
Me
N
N
NH
Me
MgBr
MgBr
MgBr
29
E-28
30
Me
C3F7
C3F7 Me
C3F7 Me
(1) H2O
(2) O2
ꢂMeOH
N
N
N
ꢁ
MgBr
Me
Me
33 (92 %)
32
31
ꢂ
HO
Scheme 4.

Acridine Synthesis
199
F7C3
MgBr
disappearance of these peaks and a gradual appearance of a peak
corresponding to acridine 33. The virtually identical mass
spectra for the intermediate products E-27 and Z-27 strongly
suggested that these molecules are isomeric. The identical
molecular compositions of these intermediate products, fully
consistent with the assigned structures, were established by fast
atom bombardment high-resolution mass spectrometry (FAB
HRMS), which also showed the Mþ1 peak as the base peak in
both cases. The crude mixture of Z-27 and E-27 solidifies at
ꢀ258C, and these compounds are stable in the solid form at
ꢀ258C for at least two weeks. At room temperature, these
compounds are oils and their decomposition can be observed
by 1H NMR. The isomer E-27 is more stable (half-life of 6 h at
238C) and was isolated in pure form by flash chromatography on
19
N
MgBr
34
35
Electrocyclization
C3F7
C3F7
Rearrangement
N
N
Mg
BrMg
1
Br
silica gel and fully characterized by H NMR. The spectrum
gave a 6-proton singlet associated with the two methyl groups,
as expected; and the aromatic proton pattern was fully consistent
with the given structure. The proton chemical shift assignment
was obtained by using COSY analysis. The E geometry was
established by using NOE experiments. Due to the inherent
instability of E-27, extensive 13C–19F couplings, and a relatively
long period of time required to accumulate 13C NMR, a quality
13C NMR spectrum of E-27 could not be obtained. Nevertheless,
the partial 13C NMR spectrum of the compound, obtained before
its decomposition became excessive, clearly showed one signal
associated with the two methyl groups. In the 19F NMR
spectrum of E-27, two CF2 moieties resonate at d 42.4 and
44.5, and the CF3 group resonates at d 79.2. This resonance
pattern is typical of a heptafluoropropyl group. The GC trace of
the mixture, respective mass spectra of Z-27 and E-27, and
1H NMR spectrum of E-27 are provided in the Supplementary
Material.
37
36
H2O
C3F7
C3F7
O2
N
N
ꢁ
ꢂ
HO
38
39
C3F7
Importantly, subjecting the isolated compound E-27 to the
reaction with 2,6-dimethylphenylmagnesium bromide (6) pro-
duced the acridine 33. The yield based on the substrate E-27
was almost quantitative. A similar treatment of the 1 : 1 mixture
of Z-27 and E-27, promptly isolated by flash chromatography,
produced acridine 33 in a yield greatly exceeding the contents of
the component E-27 in the mixture. The high yield of the product
obtained from the mixture of isomers shows that the two isomers
undergo equilibration. On the other hand, leaving the mixture of
Z-27 and E-27 or pure compound E-27 in the absence of base at
room temperature caused decomposition of the samples without
formation of acridine 33. This result demonstrates that it is the
magnesium derivative E-28, not neutral compound E-27, that
undergoes electrocyclization to the dihydroacridine system 30.
In summary, there is little doubt that the aza-dienes E-27 and
E-28 are precursors to the acridine 33. Similar intermediate
products have been suggested to be involved in base-mediated
transformations of 2-(trifluoromethyl)aniline, but were never
isolated.[5–11] The substituted aza-ortho-xylylene E-27 is the
first example of the isolated intermediate product of the many
base-mediated experiments conducted with 2-(perfluoroalkyl)
aniline.
It was suggested previously that the acridine 33 can be
formed directly by elimination of methylmagnesium bromide
from the cyclization product 30.[13] This simplistic explanation
proved to be incorrect. Our numerous trapping experiments for
methylmagnesium bromide in the reaction mixture failed to
show the presence of this organometallic reagent. First, aceto-
phenone, benzophenone, or fluorenone was allowed to react
with methylmagnesium bromide, and the corresponding adduct
was characterized by GC–MS. Then, the reaction of 19 with 6
N
40
OH
Scheme 5.
molecular oxygen generates N-methylacridinium intermediate
product 32, which is the direct precursor to the observed acridine
33. As can be seen from Scheme 4, the acridine 33 is formed in
an SN2 nucleophilic attack of hydroxide ion on the methyl group
of 32. The presumed high reactivity of 32 in nucleophilic
substitution reaction is expected because the positive charge
of the leaving group (acridine 33) and, therefore, its leaving
ability is strongly enhanced by the presence of an electron-
withdrawing perfluoroalkyl group. An important result of this
analysis is that the mysterious lack of a methyl group in the
product is due to removal of this group in the form of a methanol
molecule. This conclusion receives more attention in Scheme 5.
The given mechanism is strongly supported by the following
experimental facts. While conducting the reaction of Scheme 4,
the progress was monitored by taking small aliquots, quenching
with water, and immediately analyzing the products by gas
chromatography mass spectrometry (GC–MS). Peaks corre-
sponding to two intermediate products E-27 and Z-27 in the
ratio of 1 : 1 increased in size when the mixture was left for 1 h
at ꢀ508C, and the pattern did not change with time at ꢀ508C
thereafter. Raising the temperature to 238C resulted in slow

200
L. Strekowski et al.
was conducted as per usual, the ketone was added to the mixture,
then after sufficient time, the mixture was quenched with water
and analyzed for the presence of a ketone–methyl adduct.
Regardless of the ketone used, there was clearly no ketone–
methyl adduct in the reaction mixture. On the other hand, in a
control experiment, the mixture of 19 and 6 was allowed to react,
then treated with 0.1 equivalent of methylmagnesium bromide
followed by addition of a ketone. This treatment clearly showed
the presence of a ketone–methyl adduct in the mixture.
In the same way, an interesting experimental observation is the
regioselective electrocyclization with the involvement of the
ortho-C-Me group followed by loss of a methanol molecule for
adducts that initially contain both the ortho-C-Me and ortho-
C-H moieties. This feature can be explained in terms of
reversibility of the electrocyclization that, however, in case of
the involvement of the methyl is apparently followed by
rearrangement (C - N transfer of the methyl) to generate a
more thermodynamically stable product.
It is suggested in the mechanism shown in Scheme 4 that the
rearranged[16] magnesium derivative 31 is the final intermediate
product present in the mixture before quenching. Exposure to
water and molecular oxygen during workup is required for the
formation of the final product 33. In complete agreement with
this mechanistic pathway, the GC–MS analysis of the samples
before contact with water and molecular oxygen did not show
the presence of 33. In a similar way, the acridine 33 was absent
after treatment of the mixture with deoxygenated water only
without contact with air. Treatment of the reaction mixture with
water followed by exposure to air are necessary for the forma-
tion of the final product.
Conclusion
The structure of the product of the reaction between a
2-(perfluoroalkyl)aniline and an aromatic Grignard reagent is
dependent on the steric constrains in the intermediate products
of the type of 27 and similar aza-ortho-xylylene analogues. Low
steric hindrance facilitates the addition reaction of the Grignard
reagent with the aza-ortho-xylylene, resulting in the formation
of a tetraarylmethane or triarylmethane (molecular propellers).
Using 2-(trifluoromethyl)aniline as the substrate, addition of the
second molecule of a 2-monoalkyl- or 2,6-dialkyl-substituted
Grignard reagent is still possible, producing a 9-arylacridine as
the final product. The reaction of higher 2-(perfluoroalkyl)
anilines with substituted Grignard reagents generates the
substituted intermediate aza-ortho-xylylene of type 27, whereby
the steric conditions prevent further addition reaction, and this
sterically hindered intermediate product undergoes electro-
cyclization under basic conditions instead. The obtained acri-
dines are devoid of the ortho-alkyl group of the Grignard
reagents. The loss of the alkyl group is thoroughly explained by
the experimentally elucidated mechanism.
The loss of a small molecule of methanol in the final step
caused a major problem in the elucidation of the final mecha-
nistic pathway because this small molecule could not be identi-
fied using common experimental methods. It was thought that
the problem could be solved using Grignard reagent 34 derived
from 1-bromotetraline (Scheme 5). More specifically, the use of
this reagent would allow identification of the functionality with
which the alkyl group is eliminated from the system because the
functionalized alkyl group would be tethered to the acridine
product. Using the analogy with Scheme 4, the reaction pathway
35 - 36 - 37 - 38 - 39 - 40 could be predicted. Indeed,
the substituted acridine 40 was obtained in 58 % yield and
rigorously characterized including HRMS and elemental analy-
sis. The D2O exchangeable hydroxyl function resonates as a
Experimental
2-(Perfluoroalkyl)anilines 8, 16, 19, and 20 were synthesized
using the previously published procedure[1–4]. The Grignard
reagents 4, 6, 9, 13, and 34 were generated by the reaction of
magnesium turnings with the corresponding aryl bromides in
THF at 708C as previously described[16]. THF was distilled from
sodium benzophenone ketyl immediately before use and all
reactions were conducted under an atmosphere of nitrogen.
Flasks were fitted with rubber septa and Teflon-coated magnetic
stirring bars were used. Thin layer chromatography (TLC) was
carried out using silica gel with a fluorescent indicator (Kodak
13181). Spots were visualized by inspection under UV light.
Crude mixtures were analyzed and mass spectra of pure com-
ponents were obtained on a Shimadzu GC instrument coupled
with an electron impact (EI) mass spectrometer. Preparative
chromatography was performed using EM Reagents silica gel.
Melting points (Pyrex capillary) are uncorrected. 1H NMR
spectra (300 MHz), 13C NMR spectra (75 MHz), and 19F NMR
spectra (298 MHz, C6F6 as internal standard) were recorded on a
Varian Unity 300 spectrometer for solutions in CDCl3 (free
bases) and [D6]DMSO (HBr salts). High-resolution mass
spectra (FAB HRMS) were recorded on a VG Analytical 70-SE
spectrometer.
1
broad peak at d 8.56 in the H NMR spectrum recorded in
CDCl3. The 19F NMR spectrum of 40 shows the presence of a
1
perfluoropropyl group and both the H NMR and 13C NMR
spectra indicate the presence of a tetramethylene substituent in
the molecule.
Finally, we wish to comment on the formation of 9-arylacri-
dines[13] 5 and 7 by the treatment of 2-(trifluoromethyl)aniline
(1) with the respective Grignard reagents 4 and 6 (Scheme 1). In
the light of the mechanistic results obtained as part of this work,
the previously suggested mechanism for 5 and 7 is incorrect. The
formation of these compounds is completely consistent with a
unified mechanism that can be derived from the mechanism of
Scheme 4 by taking into account the different structures of
substrates 1 and 19 and the resulting differences in steric
hindrance and steric condition-driven reactivity of the interme-
diate products. Thus, the sterically congested intermediate
product 27 does not undergo addition of the Grignard reagent
6, and its magnesium derivative 28 undergoes slow electrocy-
clization to a dihydroacridine. In contrast, the corresponding
analogue of 27 derived from 1 (F instead of C3F7) is consider-
ably less sterically hindered and, as such, can undergo addition
reaction with 6 to generate a diaryl adduct. Subsequent electro-
cyclization followed by a sequence of reactions, analogous to
those shown in Scheme 4 (rearrangement and oxidation fol-
lowed by the loss of a methanol molecule), are suggested to yield
the final 9-arylacridine 7.
Reactions of 2-(Perfluoroalkyl)anilines with
Arylmagnesium Bromides
A solution of Grignard reagent (10 mmol) in THF (10 mL)
was stirred at ꢀ708C and treated dropwise with a solution of
2-(perfluoroalkyl)aniline (3 mmol) in THF (10 mL). The
resulting brown mixture was allowed to reach 238C within 1 h
and stirred at 238C for several hours until TLC and GC–MS
An important experimental result is the lack of electrocycli-
zation of the adducts derived from phenylmagnesium bromide.

Acridine Synthesis
201
analyses showed the absence of the starting aniline. The mixture
was quenched with water (5 mL), purged with air, and con-
centrated to 15 mL on a rotary evaporator. The residue was
extracted with ether (3 ꢁ 20 mL), dried, and concentrated.
Product was purified by radial chromatography on silica gel
eluting with pentanes or pentane/ether (10 : 1).
(15 %, Mþ–CH3), 275 (100 %, Mþ). Found: C 69.88, H 4.56,
N 5.03 %; (Mþ) 275.0925. Anal. Calc. for C16H12F3N: C 69.81,
H 4.39, N 5.09 %; (Mþ) 275.0922.
9-(Pentafluoroethyl)acridine (17)
This compound was obtained from 16 and 9 as a light yellow
oil (93 % yield). dH 7.65 (2H, t, J 8), 7.82 (2H, t, J 8), 8.35 (4H,
m). dF 62.6 (2F), 79.2 (3F). m/z (EI) 228 (100 %, Mþ–CF3),
297 (45 %, Mþ). m/z (HRMS FAB) 297.0567; Mþ (C15H8NF5)
requires 297.0577.
Isolation and Purification
Perfluoroalkyl-substituted acridines are highly volatile, and it is
usual to lose most of the product during careless workup. Low-
boiling solvents, such as pentane and diethyl ether, should be
used for workup including chromatography and crystallization
steps. The preferred purification is silica gel chromatography
eluting with pentanes or pentanes/ether (10 : 1), followed by
partial removal of the solvent, and then treatment of the solution
with hydrobromic acid in diethyl ether to obtain a precipitate of
the hydrobromide.
17ꢂHBr
Yield 95 %, mp 186–1878C. dH 7.66 (2H, m), 7.83 (2H, m),
8.35 (4H, m). Partial dC 125.0, 128.1, 129.7, 130.8, 148.9. Anal.
Calc. for C15H9BrNF5: C 47.64, H 2.40, N 3.70. Found: C 47.30,
H 2.65; N 3.64 %.
In a typical experiment, a solution of an acridine (1 mmol) in
diethyl ether (10 mL) was stirred, treated dropwisewitha solution
of hydrobromic acid (48 %) and ether (1: 10, 5 mL), and then
cooled. The resulting precipitate was crystallized from ether.
1,3-Dimethyl-9-(pentafluoroethyl)acridine (18)
This compound was obtained from 16 and 13 as a light
yellow oil (67 % yield). dH 2.46 (3H, s, 3-CH3), 2.63 (3H, t, J 4,
1-CH3), 7.26 (1H, s), 7.51 (1H, t, J 8), 7.68 (1H, t, J 8), 7.79 (1H,
s), 8.16 (2H, m). dF 74.4 (2F), 90.2 (3F). m/z (EI) 256 (100 %,
Mþ–CF3), 325 (80 %, Mþ). Found: C 62.91, H 3.65, N 4.23 %;
(Mþ) 325.0880. Anal. Calc. for C17H12 F5N: C 62.77, H 3.72,
N 4.31 %; (Mþ) 325.0890.
9-(Trifluoromethyl)acridine (10)
This compound was obtained from 8 and 4 (46 % yield), and
from 8 and 9 (47 % yield). dH 7.68 (2H, t, J 8), 7.84 (2H, t, J 8),
8.32 (2H, d, J 9), 8.49 (1H, d, J 9). dF 111. m/z (EI) 197 (20 %,
Mþ–CF2), 228 (5 %, Mþ–F), 247 (100 %, Mþ). Found: C 68.15,
H 3.14, N 5.60 %; (Mþ) 247.0592. Anal. Calc. for C14H8 F3N:
C 68.02, H 3.26, N 5.67 %; (Mþ) 247.0609.
3-Methyl-9-(heptafluoropropyl)acridine (22)
This compound was obtained from 19 and 21 as a yellow oil
(86 % yield). m/z (EI) 242 (100 %, Mþ–C2F5), 361 (30 %,
Mþ). m/z (HRMS FAB) 361.0717; Mþ (C17H10NF7) requires
361.0702.
[1,1-Bis-(2-ethylphenyl)-2,2,2-trifluoroethyl]aniline (11)
This compound was obtained from 8 and 4 as a brown oil
(46 % yield). dH 1.09 (6H, m), 2.47 (2H, q, J 8, CH2), 2.71 (2H, q,
J 8, CH2) at 238C, 2.59 (4H, q, J 8, 2CH2) at the coalescence
temperature of 558C, 6.33–8.41 (12H). dF (608C) 28.1. m/z (EI)
383 (100 %, Mþ). Found: C 75.00, H 6.48, N 3.59 %; (Mþ)
383.1868. Anal. Calc. for C24H24 F3N: C 75.17, H 6.31, N
3.65 %; (Mþ) 383.1861.
22ꢂHBr
Yellow crystals (96 % yield), mp 212–2138C [from ether/
pentane (2 : 1)]. dH 2.78 (3H, s, CH3), 7.78 (1H, d, J 9), 7.93
(1H, t, J 8), 8.19 (1H, t, J 8), 8.44 (1H, d, J 9), 8.53 (1H, d, J 9),
9.17 (1H, s), 9.47 (1H, d, J 9). dF 38.9 (2F), 66.0 (2F), 82.3 (3F).
Partial dC 22.7, 99.9, 119.9, 121.9, 129.8, 133.4, 136.1, 140.5,
149.6. Anal. Calc. for C17H11 F7N: C 46.18, H 2.51, N 3.17.
Found: C 46.59, H 2.61, N, 3.08 %.
2-[1,1-Bis-(2-tolyl)-2,2,2-trifluoroethyl]aniline (12)
This compound was obtained from 8 and 4 as a brown oil
(44 % yield). dH 2.02 (6H, s, CH3), 2.06 (6H, s, CH3) at 238C,
2.04 (6H, s) at the coalescence temperature of 508C, 3.23 (2H, br
s, exchangeable with D2O), 6.33–8.41(12H). dF (608C) 112. m/z
(GC) 356 (100 %, Mþþ1). Found: C 74.48, H 5.81, N 4.03 %;
(Mþ) 355.1550. Anal. Calc. for C22H20 F3N: C 74.35, H 5.67,
N 3.94 %; (Mþ) 355.1548.
1,3-Dimethyl-9-(undecafluoropentyl)acridine (23)
This compound was obtained from 20 and 13 as a yellow oil
(55 % yield). dH 2.53 (3H, s), 2.71 (3H, t, J 5), 7.36 (1H, s), 7.57
(1H, t, J 7), 7.74 (1H, t, J 7), 7.86 (1H, s), 8.20 (1H, d, J 8), 8.34
(1H, d, J 8). dF 35.9 (2F), 39.8 (2F), 50.2 (2F), 66.0 (2F), 81.1
(3F, s). Partial dC 21.2, 24.7, 124.1, 124.4, 125.2, 125.3, 125.4,
127.1, 127.2, 129.1, 129.5, 129.8, 135.4, 139.4, 147.8, 150.2.
m/z (EI) 256 (100 %, Mþ–C4F9), 456 (10 %, Mþ–F), 475 (45 %,
Mþ). Found: C 50.25, H 2.61, N 2.80 %; (Mþ) 475.0808. Anal.
Calc. for C20H12 F11N: C 50.54, H 2.54, N 2.95 %; (Mþ)
475.0794.
1-Methyl-9-(trifluoromethyl)acridine (14)
This compound was obtained from 8 and 6 as a light yellow
oil (74 % yield). m/z (EI) 240 (10 %, Mþ–F), 261 (100 %, Mþ).
Found: C 68.70, H 3.80, N 5.30 %; (Mþ) 261.0772. Anal. Calc.
for C15H10 F3N: C 68.96, H 3.86, N 5.36 %; (Mþ) 261.0765.
1-Methyl-9-(heptafluoropropyl)acridine (33)
1,3-Dimethyl-9-(trifluoromethyl)acridine (15)
This compound was obtained from 19 and 6 as a yellow oil
(92 % yield). dH 2.70 (3H, t, J 5), 7.52 (1H, d, J 7), 7.64 (2H, m),
7.78 (1H, t, J 7), 8.09 (1H, d, J 8) 8.25 (1H, d, J 9), 8.37 (1H, d,
J 9). dF 46.7 (2F), 78.9 (2F), 82.0 (3F). Partial dC 24.8, 125.2,
127.5, 128.6, 129.1, 129.5, 130.1, 132.1, 133.6, 147.9, 149.9.
m/z (EI) 242 (100 %, Mþ–C2F5), 361 (45 %, Mþ). m/z (HRMS
ESI) 362.0793; Mþþ1 (C17H11NF7) requires 362.0780.
This compound was obtained from 8 and 13 as light yellow
crystals (80 % yield), mp 101–1028C. dH 2.54 (3H, d, J 10), 2.70
(3H, t, J 3), 7.31 (1H, d, J 9), 7.57 (1H, t, J 7), 7.75 (1H, d, J 8),
7.79 (2H, m), 8.21 (1H, t, J 9), 8.31 (1H, s). dF 113. Partial dC
21.2, 24.6, 124.0, 124.4, 125.2, 125.3, 125.3, 127.0, 127.4,
129.4, 130.0, 133.0, 135.3, 139.3, 148.0, 150.4. m/z (EI) 260

202
L. Strekowski et al.
33ꢂHBr
Yield 98 %, mp .2208C. Anal. Calc. for C17H11F7N:
C 46.18, H 2.51, N 3.17. Found: C 46.49, H 2.61, N 3.08 %.
Synthesis of Acridine 33 Starting with
Intermediate Product E-27
The GC–MS analysis of the reaction mixture discussed above
indicated that the two intermediate products E-27 and Z-27 (and
their magnesium derivatives E-28 and Z-28) are stable at ꢀ708C
and the acridine system is not generated at this temperature
even after several hours. Raising the temperature to 238C for 2 h
followed by usual workup resulted in the formation of acridine
33 in 92 % yield. On the other hand, treatment of the pure
intermediate product E-27 with 2,6-dimethylphenylmagnesium
bromide (6) using a general procedure afforded product 33 in a
virtually quantitative yield.
4-(4-Hydroxybutyl)-9-(heptafluoropropyl)acridine (40)
This compound was obtained from 34 and 19 as a light
yellow oil (58 % yield). dH 1.80 (2H, m, CH2), 1.95 (2H, m,
CH2), 3.45 (2H, t, J 7.5, acridine-CH2), 4.14 (2H, t, J 6.8,
O-CH2), 7.53 (1H, m, CH), 7.61 (2H, m, CH), 7.76 (1H, m, CH),
8.21 (1H, m, CH), 8.33 (2H, m, CH), 8.56 (1H, br s, exchange-
able with D2O, OH). Partial dC 27.1, 27.5, 31.9, 76.7, 123.1,
124.8, 127.9, 128.0, 128.6, 129.3, 131.3. dF 38.3 (2F), 66.0 (2F),
82.0 (3F). m/z (EI) 419 (100 %, Mþ). Found: C 55.34, H 3.86, N
2.98 %; (Mþþ1) 420.1212. Anal. Calc. for C20H16F7NO: C
55.50, H 4.08, N 3.23 %; (Mþþ1) 420.1198.
Supplementary Material
The Supplementary Material contains EI mass spectra of E-27
and Z-27, and is available on the Journal’s website. The spec-
trum of Z-27 was obtained by analysis of the crude mixture using
GC–MS immediately after quenching. The spectra show the
same sets of major peaks (abundance .5 %), with the intensity
differences of the corresponding peaks in the two spectra no
greater than 10 %.
Isolation of E-6-[1-(2,6-dimethylphenyl)-2,2,3,3,4,4,4-
heptafluorobutylidene]-2,4-cyclohexadien-1-imine (E-27)
A
solution of 2,6-dimethylphenylmagnesium bromide
(6, 3 mmol) in THF (3mL) was stirred at ꢀ788C and treated
dropwise with solution of 2-(nonafluorobutyl)aniline
a
(19, 310 mg, 1 mmol) in THF (3mL). The resultant brown
mixture was kept at ꢀ708C until GC–MS analysis showed the
absence of 19. The GC–MS analysis was conducted on a
Shimadzu GC-17A chromatograph coupled to a QP-5000 mass
selective detector, using a 15m ꢁ 0.25mm XTI-5 column coated
with poly(dimethylsiloxane). Two well-separated major products
were observed (ratio 1 : 1) by GC–MS, and they produced virtu-
ally identical mass spectra (see Supplementary Material). Product
E-27 showed a retention time of 6.59 min and its isomer Z-27
appeared after 6.64 min. The mixture was quenched with cold
water (3mL) and concentrated to 5 mL without heating on a
rotary evaporator. The residue was extracted with cold ether
(3ꢁ 10 mL) and then quickly concentrated without heating.
A mixture of E-27 and Z-27 are stable at ꢀ258C, but attempted
separation by radial chromatography on silica gel eluting with
pentane afforded pure product E-27 only, which was eluted first,
followed by impure isomer Z-27. The separated isomer E-27 was
kept in solid state at ꢀ258C. As shown by 1H NMR, compound
E-27 is stable in the solid state at ꢀ258C for several days. Due to
its instability, the isomer Z-27 could not be obtained in pure state.
Compound E-27 was characterized as follows: an oil at 238C; an
amorphous solid at ꢀ258C; 45% yield; dH 7.13 (1H, t, J 7,
H4),7.16 (1H, t, J 7, H40), 7.07(2H, 2d, J 7, H30 and H50 ofthe 2,6-
diMePh), 6.76 (2H, 2d, J 8, H2and H5), 6.66 (1H, t, J 8, 4-H), 3.94
(1H, s, exchangeable with D2O, NH), 2.26 (6H, s, 2ꢁCH3) – these
chemical shift assignments were obtained using COSY. 1H NMR
NOE: the observed through-space interaction of the methyl groups
with the imino proton is consistent with the E-configuration,
irradiation at 3.94 (NH) gave signals at d 2.26 (2ꢁCH3) and 6.76
(H2), irradiation at d 2.26 (2ꢁCH3) gave a signal at d 3.94 (NH)
only; dF 79.2 (3F), 44.5 (2F), 42.4 (2F); m/z (EI) 339 (20 %,
Mþ–2F), 359 (100 %, Mþ–FþH), 377 (5 %, Mþ). m/z (HRMS
ESI) 378.1085; MþþH (C18H15F7N) requires 378.1093.
References
[1] T. Fuchikami, I. Ojima, J. Fluor. Chem. 1983, 22, 541. doi:10.1016/
S0022-1139(00)81229-4
[2] N. Yoshino, M. Kitamura, T. Seto, Y. Shibata, M. Abe, K. Ogino, Bull.
Chem. Soc. Jpn. 1992, 65, 2141. doi:10.1246/BCSJ.65.2141
[3] M. Tordeux, B. Langlois, C. Wakselman, J. Chem. Soc., Perkin Trans.
1 1990, 2293. doi:10.1039/P19900002293
[4] L. Strekowski, M. Hojjat, S. E. Patterson, A. S. Kiselyov, J. Hetero-
cycl. Chem. 1994, 31, 1413. doi:10.1002/JHET.5570310619
[5] L. Strekowski, A. S. Kiselyov, Trends Heterocycl. Chem. 1993, 3, 73.
[6] A. S. Kiselyov, L. Strekowski, Org. Prep. Proced. Int. 1996, 28, 289.
doi:10.1080/00304949609356536
[7] L. Strekowski, J. Zhang, E. Paliakov, M. Say, Recent Res. Dev. Org.
Chem. 2004, 8, 1.
[8] L. Strekowski, E. Wolinska, M. Mojzych, in Synthetic and Biophysical
Studies of DNA Binding Compounds (Eds M. Lee, L. Strekowski)
2007, pp. 263–278 (Transworld Research Network: Trivandrum).
[9] A. S. Kiselyov, Mini-Rev. Org. Chem. 2007, 4, 183. doi:10.2174/
157019307780599289
[10] K. Wojciechowski, Eur. J. Org. Chem. 2001, 3587. doi:10.1002/1099-
0690(200110)2001:19,3587::AID-EJOC3587.3.0.CO;2-5
[11] Y. Kobayashi, I. Kumadaki, Acc. Chem. Res. 1978, 11, 197.
doi:10.1021/AR50125A004
[12] L. Strekowski, S.-Y. Lin, H. Lee, J. C. Mason, Tetrahedron Lett. 1996,
37, 4655. doi:10.1016/0040-4039(96)00913-6
[13] J. Zhang, J. Sa˛czewski, E. Wolin´ska, L. Strekowski, Heterocycl.
Commun. 2013, 19, 245.
[14] J. Zhang, J. Sa˛czewski, L. Strekowski, Heterocycl. Commun. 2013,
19, 357.
[15] L. Strekowski, H. Lee, S.-Y. Lin, A. Czarny, D. Van Derveer, J. Org.
Chem. 2000, 65, 7703. doi:10.1021/JO001001C
[16] T. Holm, I. Crossland, in Grignard Reagents: New Developments
(Ed. H. G. Richey, Jr) 2000, Ch. 1, pp. 1–26 (John Wiley and Sons:
Chichester, West Sussex).