Nuclear fluorination of 3,5-diarylisoxazoles with Selectfluor

Article abstract of DOI:10.1016/j.jfluchem.2004.08.001

C-4 Fluorination of a series of 3,5-diarylisoxazoles has been accomplished using the N-F reagent Selectfluor. With substrates containing neutral or activating substituents on the 5-phenyl ring, acetonitrile at room temperature or at reflux coul

Full text of DOI:10.1016/j.jfluchem.2004.08.001

Journal of Fluorine Chemistry 125 (2004) 1939–1945
www.elsevier.com/locate/fluor
Nuclear fluorination of 3,5-diarylisoxazoles with Selectfluor1
Chad E. Stephens*, Jacqueline A. Blake
Department of Chemistry, Center for Biotechnology and Drug Design, Georgia State University, Atlanta, GA 30303, USA
Received 27 July 2004; received in revised form 3 August 2004; accepted 4 August 2004
Available online 18 September 2004
Abstract
C-4 Fluorination of a series of 3,5-diarylisoxazoles has been accomplished using the N-F reagent Selectfluor1. With substrates
containing neutral or activating substituents on the 5-phenyl ring, acetonitrile at room temperature or at reflux could be used as solvent.
However, when deactivating substituents were present, a higher reaction temperature was required for which sulfolane was found to be a
good solvent. At this higher temperature, a unique trifluorination of the isoxazole nucleus by an addition mechanism occurred as a side
reaction.
# 2004 Elsevier B.V. All rights reserved.
Keywords: Heterocycle; Isoxazole; Fluorination; Selectfluor1; Sulfolane; Substituent effects
1. Introduction
to explore their use for the fluorination of our isoxazole
system. Herein, we describe our work in this area leading
to convenient procedures for the selective C-4 fluorination
of 3,5-diarylisoxazoles. In addition, we also describe the
isolation of a trifluorinated by-product from some of our
higher-temperature reactions, which was formed by an
addition reaction to the isoxazole nucleus following the
initial substitution.
3,5-Diarylisoxazoles have been studied recently as
potential anti-cancer [1] and anti-microbial agents [2], as
potential estrogen receptor modulators [3], as liquid crystals
[4], and as linear dyes with application to light-harvesting
systems [5]. Previously, we have described the convenient
C-4 halogenation (Cl, Br, or I) of a series of these
compounds, using N-halosuccinimides [6]. In that study,
the effect of the substituent on the 5-phenyl ring was found
to exert a profound influence on the reactivity of the
isoxazole.
2. Results and discussion
As fluorinated compounds often possess unique biolo-
gical [7] and physiochemical [8] properties compared to the
parent systems, we have become interested in the fluorina-
tion of these and other heterocycles. Our literature search
revealed that while the synthesis of 4-fluoro-3,5-dipheny-
lisoxazole via a ring-closure process has been described [9],
there has been no report of the direct C-4 fluorination of
3,5-diarylisoxazoles. In fact, there appears to have been no
report of the direct fluorination of any isoxazole at any ring
position. As a number of easily handled N-F fluorinating
reagents are now readily available [10], we decided
The N-F fluorinating reagents used in this study are
shown in Fig. 1. Initial experiments were carried out with
the parent isoxazole 4a and Selectfluor1 (1) (1.1 eq.)
(Scheme 1). Although no reaction occurred in acetonitrile at
room temperature after several hours, a reaction did occur at
reflux. However, although the desired 4-fluoroisoxazole was
formed quite selectively, consumption of the starting
material was always incomplete (ꢀ30–50% conversion by
TLC analysis) even upon extended reaction times (>24 h).
A similar outcome was observed when nitromethane was
used as solvent, while no fluorination occurred when acetic
acid or methanol was used. Attempts to drive the reaction
that occurred in acetonitrile by including additives such as
* Corresponding author. Tel.: +1 404 651 3868; fax: +1 404 651 1416.
E-mail address: checes@panther.gsu.edu (C.E. Stephens).
0022-1139/$ – see front matter # 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.jfluchem.2004.08.001

1940
C.E. Stephens, J.A. Blake / Journal of Fluorine Chemistry 125 (2004) 1939–1945
1
characterized by H, 13C, and 19F NMR, as well as by IR,
MS and combustion analysis (Tables 1 and 2).
The 19F NMR spectrum of 5a revealed one up-field
singlet (ꢁ177.9 ppm), which is consistent with that
previously reported [9], while the 1H NMR lacked the
characteristic singlet for the C-4 proton of the starting
isoxazole. In the 13C NMR, a significant downfield shift was
observed for C-4 of the fluorinated isoxazole (140.9 ppm)
Fig. 1. Structures of N-F reagents tested.
compared to that of the parent isoxazole (98.6 ppm). The 13
C
trifluoroacetic acid [11], nitrobenzene [12], KHCO3, or
NaOTf [10e] were unsuccessful. The use of excess
fluorinating reagent (2 eq.) also afforded no apparent
increase in conversion. We thus, speculate that a competitive
N-fluorination of the isoxazole nitrogen may be the reason
for the incomplete reaction, a possibility that has been put
forth by other authors [13] with regard to the fluorination of
nitrogen heterocycles. Indeed, a trace amount of a polar
product that may have been the N-F isoxazole was often
observed by TLC, however attempts to isolate this
compound were unsuccessful. Hydrolysis of the N-F adduct
on TLC or upon work-up would regenerate the starting
isoxazole.
NMR also showed the expected 13C-19F coupling, with C-4
appearing as a doublet with a large one-bond coupling (1JC-
F = 256 Hz), and C-3 and C-5 appearing both as doublets
with smaller two-bond couplings (2JC-F = 10.9 and
18.9 Hz). Three- and four-bond 13C-19F couplings to each
of the phenyl rings were also observed. Finally, the mass
spectrum was noticeably similar to that of the parent
isoxazole (4a) [14] in that the most intense signals were
those corresponding to the 5-PhCO+ (M+ ꢁ 134) and 5-Ph+
(M+ ꢁ 162) ions. Fragments resulting from loss of HF, CO,
and the FCCO radical were also observed.
With the parent 4-fluoroisoxazole in hand, we next turned
to the fluorination of the substituted isoxazoles 4b-e.
Reaction of the methyl-substituted isoxazole 4b with
Selectfluor1 proceeded as did the parent, giving 5b in
34% yield. Fluorination of the activated methoxy substituted
analogue 4c, in contrast, led to a complex mixture (TLC
analysis) consisting of the 4-fluoroisoxazole 5c, starting
material, and several other unidentified products possibly
arising from non-selective phenyl fluorination and/or
demethylation [15]. The reaction was much more selective
at room temperature, however, yielding only the 4-
fluoroisoxazole as a new product. Separation from remain-
ing starting material gave the desired product 5c in 28%
yield. As the parent isoxazole did not react with Selectfluor1
at room temperature, the methoxy-substituted analogue is
clearly activated.
In a continued effort to improve the reaction, other N-F
fluorinating reagents were explored. No reaction occurred
when parent isoxazole 4a was treated with N-fluoropyr-
idinium triflate (2a) or with N-fluorobenzenesulfonimide (3)
(1.1 eq., refluxing acetonitrile). However, fluorination did
take place when the more reactive N-fluoro-2,6-dichlor-
opyridinium triflate (2b) was used, although an incomplete
conversion was again observed which offered no apparent
improvement compared to the less expensive Selectfluor1.
In the end, the use of Selectfluor1 in refluxing
acetonitrile without any additive was judged to be the best
of the conditions explored for fluorinating isoxazole 4a.
Despite the incomplete conversion, we were still pleased to
obtain 4-fluoroisoxazole 5a in 32% yield, using these
conditions following silica gel chromatography, to remove
residual starting material. The compound was fully
Fluorination of the deactivated bromo- and trifluoro-
methyl-substituted isoxazoles 4d and 4e was also explored.
Scheme 1.

C.E. Stephens, J.A. Blake / Journal of Fluorine Chemistry 125 (2004) 1939–1945
1941
Consistent with our previous work with this series of
isoxazoles [6], the reaction of either these compounds with
Selectfluor1 in refluxing acetonitrile was quite sluggish,
giving only traces of the desired 4-fluoroisoxazoles. As a
much higher reaction temperature appeared needed, we
decided to explore the use of higher boiling solvents. A scan
of the literature revealed that the use of higher reaction
temperatures for fluorination of aromatics with N-F reagents
has been quite limited. We were most encouraged by a report
of the fluorination of a nitro-substituted quinoline with NFSI
(3) at 130 8C, although this reaction was performed without
solvent [16]. The only examples that we found related to
higher boiling solvents involve the use of the relatively non-
polar 1,1,2-trichloroethane at reflux (bp 110–115), which
has led to varied results [10c,16]. As we felt the polar
Selectfluor1 would perform best in solution, we opted to
explore a few of the common polar aprotic solvents
[dimethylformamide (DMF), dimethylsulfoxide (DMSO),
sulfolane, dimethylsulfone] for our fluorination.
To our satisfaction, a significant amount of fluorination
took place when isoxazole 4d was reacted with Selectfluor1
at 120 8C for 3–4 h in three-fourth of these solvents, with the
notable exception being DMSO. As no new product was
formed when DMSO was used, it appears that Selectfluor1
may react with this solvent (no report of the reaction of non-
deprotonated sulfoxides with Selectfluor1 could be found.
However, a reaction with sulfides is known to occur [10f].
The reaction with sulfimides is also known [17]). Of the
other three solvents, sulfolane emerged as our solvent of
choice, as the use of dimethylsulfone was complicated by
sublimation and reactions in DMF became much darker in
color. While sulfolane has been much utilized for
fluorination via halex substitution reactions [18], we could
not find any report of it being used previously with N-F
reagents.
Utilizing the higher reaction temperature, 4-fluoroisox-
azoles 5d and 5e were obtained in 39% and 30% yield,
respectively. Again, conversion was incomplete and a
significant amount of starting material remained unchanged
in each case. Interestingly, a small amount (10%) of a more
lipophilic by-product was also isolated from these high-
temperature reactions which was determined to be the 4,4,5-
trifluoroisoxazoline 6 (Scheme 1) based largely on the
multiplicity observed in the 19F and 13C NMR (Table 2), as
well as the mass spectrum (Table 1). For example, the 19F
NMR spectrum of 6d showed three double doublets (dd),
two with both a large geminal coupling (2JF-F = 273 Hz) and
a smaller vicinal coupling (3JF-F = 3.7 or 8.6 Hz), and one
with only two vicinal couplings (3JF-F = 3.7 and 8.6 Hz).
The key signal in the 13C NMR, among others, is the triple
doublet (ddd) observed for C-4 of the isoxazoline with 1JC-F
1
2
= 274 Hz, JC-F0 = 256 Hz, and JC-F = 34.3 Hz. Finally,
proposed structures for the two most prevalent ions observed
in the mass spectrum of 6d are shown in Fig. 2. The
formation of azirines (as shown in path b) in the mass
spectrum of isoxazoles is well established [14].

Table 2
IR, 1H, 13C, and 19F NMR data for compounds 5a–e and 6d–e
Product
IR (KBr), cmꢁ1
1H NMRa, d, J (Hz)
13C NMR (DMSO-d6) d, J (Hz)
19F NMRa, d, J (Hz)
ꢁ177.9 (s)
5a
3067, 1650, 1467, 1448, 1422,
1215, 931, 774, 693, 480
7.54–7.65 (m, 6H), 7.83–7.89 (m, 4H)
124.9 (d, J = 4.6), 125.0 (d, J = 4.6), 125.8 (d, J = 3.5),
126.8 (d, J = 2.9), 129.4, 129.6, 130.8, 131.0, 140.9 (d, J = 256),
152.8 (d, J = 10.9), 153.1 (d, J = 18.9)
5b
5c
5d
5e
3073, 3039, 2916, 1649,
2.37 (s, 3H), 7.40 (d, J = 7.8, 2H),
7.58 (m, 3H), 7.71 (d, J = 8.1, 2H),
7.84 (m, 2H)
21.0, 122.1 (d, J = 5.2), 124.9 (d, J = 4.6), 125.9 (d, J = 4.0),
126.8 (d, J = 3.5), 129.3, 130.0, 130.8, 140.4 (d, J = 256), 140.7,
152.6 (d, J = 10.9), 153.3 (d, J = 19.6)
ꢁ178.6 (s)
1469, 1421, 1215, 931, 818, 691, 491
3060, 3006, 2940, 2843, 1647, 1463,
1421, 1254, 1176, 1033, 829, 689
3.83 (s, 3H), 7.16 (d, J = 9.0, 2H),
7.58 (m, 3H), 7.78 (d, J = 9.0, 2H),
7.85 (m, 2H)
55.4, 115.0, 117.5 (d, J = 5.2), 126.0 (d, J = 4.0), 126.8 (m, 2C),
129.4, 130.9, 139.8 (d, J = 255), 152.6 (d, J = 10.3),
153.3 (d, J = 19.4), 160.9
ꢁ180.0 (s)
3100, 3066, 3034, 1647, 1468, 1399,
1215, 931, 829, 818, 687, 488
7.59 (m, 3H), 7.75 (d, J = 8.4, 2H),
7.82 (d, J = 8.7, 2H),
124.0 (m, 2C), 125.7, 126.8 (d, J = 3.0), 126.9 (d, J = 5.2),
129.3, 131.0, 132.6, 141.0 (d, J = 258), 152.2 (d, J = 18.8),
152.9 (d, J = 11.3)
ꢁ176.7 (s)
7.83–7.87 (m, 2H)
3108, 3065, 1652, 1617, 1471, 1461,
1325, 1162, 1112, 1070, 845, 697, 498
7.60 (m, 3H), 7.86 (m, 2H),
7.97 (d, J = 8.7, 2H),
123.6 (q, J = 273), 125.4 (d, J = 3.5), 125.7 (d, J = 5.1),
126.2 (q, J = 3.5), 126.7 (d, J = 3.5), 128.3 (unresolved m),
129.2, 130.2 (q, J = 31.5), 130.9, 141.7 (d, J = 259),
151.6 (d, J = 19.4), 153.0 (d, J = 10.9)
ꢁ175.2 (s), ꢁ61.8 (s)
8.03 (d, J = 8.4, 2H)
6d
6e
3091, 3068, 1593, 1295, 1252, 1133,
1047, 887, 827, 683, 649
7.49–7.58 (m, 5H),
7.66 (d, J = 9.0, 2H),
7.87 (d, J = 7.2, 2H)
110.8 (ddd, J = 242, 36.3, 21.1), 123.6 (d, J = 1.9),
124.6 (ddd, J = 274, 256, 34.3), 126.0 (d, J = 2.2),
127.1 (d, J = 1.7), 127.1 (d, J = 27.9), 128.7 (dd, J = 5.1, 1.4),
129.3, 132.0, 132.4, 154.6 (dd, J = 25.5, 23.5)
111.5 (ddd, J = 242, 33.3, 18.0), 123.5 (d, J = 1.8), 123.6 (q, J = 273),
124.8 (ddd, J = 269, 257, 33.9) 125.7 (q, J = 3.4), 127.2 (d, J = 1.5),
127.7 (dd, J = 4.7, 2.0), 129.3, 131.9 (d, J = 28.1), 132.5,
133.3 (q, J = 31.5),
ꢁ124.4 (dd, J = 273, 4.3),
ꢁ117.0 (dd, J = 8.0, 4.3),
ꢁ102.3 (dd, J = 273, 8.0)
3116, 3064, 1622, 1450, 1414, 1328,
1265, 1140, 1068, 887, 837, 651
7.50–7.62 (m, 3H), 7.79 (s, 4H),
ꢁ123.8 (dd, J = 273, 3.7),
ꢁ117.3 (dd, J = 8.6, 3.7),
ꢁ102.4 (dd, J = 273, 8.6),
ꢁ64.2 (s)
7.89 (d, J = 7.5, 2H)
154.5 (dd, J = 25.3, 23.6)
a
DMSO-d6 used as solvent for 5a–e; CDCl3 used as solvent for 6d–e.

C.E. Stephens, J.A. Blake / Journal of Fluorine Chemistry 125 (2004) 1939–1945
1943
Fig. 2. Proposed structures for most prevalent ions in mass spectrum of 6d.
triflate-containing reagent 2b was used indicates that the
fluorine can also come from the N-F cation of Selectfluor1
or, more likely, from the HF that is present as a
decomposition product (the MSDS for Selectfluor1
indicates that HF is a decomposition product. We thank
the reviewer for bringing this to our attention). Finally, we
should note that a similar addition reaction has been
observed upon chlorination of such isoxazoles with HCl(aq)/
H2O2 [20]. In that case, a 4,4-dichloro-5-hydroxyisoxazo-
line was obtained, with H2O apparently serving as the
nucleophile.
Scheme 2.
The unique trifluoroisoxazoline 6 was apparently formed
by a formal addition of fluorine to the 4-fluoroisoxazole
subsequent to its formation. This is supported by the fact that
the trifluoro product 6d was also obtained (49% yield) when
4-fluoroisoxazole 5d was reacted similarly with Select-
fluor1 (2.1 molar eq. in this case) (Scheme 2).
To further explore this interesting reaction, 4-chloroisox-
azole 7 [6] was also reacted with Selectfluor1 (2.1 molar
eq.) at high temperature (Scheme 3). As determined by GC–
MS and/or 19F NMR, the main products from this reaction,
in addition to starting material, were the two 4-chloro-4,5-
difluoroisoxazoline diastereomers (8a/8b) and, unexpect-
edly, the 4,4,5-trifluoro derivative 6d. This latter compound
was apparently formed following an initial ipso substitution
of the chlorine by fluorine to give 4-fluoroisoxazole 5d.
Indeed, a trace amount of 5d was confirmed to be present in
the crude reaction mixture, along with several other
unidentified trace products. Upon purification by column
chromatography, a mixture of 8a/8b and 6d was obtained
which could not be further separated. The mixture was
characterized by 19F NMR (Section 4).
To the best of our knowledge, such an addition to an
aromatic ring upon reaction with an N-F reagent has not
previously been described, although a similar addition has
been reported to occur as a side reaction upon treating some
glycols with Selectfluor1 and a nucleophile [19]. Some of
those authors [19b] suggest that fluoride anion released from
the tetrafluoroborate counteranion serves as an opportunistic
nucleophile, adding to the substrate following the electro-
philic addition of fluorine. While such a mechanism would
appear plausible in our case, the fact that we also observed a
trace of the trifluorinated product by GC–MS when the
Scheme 3.

1944
C.E. Stephens, J.A. Blake / Journal of Fluorine Chemistry 125 (2004) 1939–1945
3. Conclusion
in an oil bath set at 120 8C for 3 h (for 4d) or 4 h (for 4e). The
pale yellow solution was then diluted with excess water to
give a white solid, which was collected and rinsed with
water. Column chromatography eluting with pentane gave
pure compound 6. Further elution with 2% ether in pentane
gave pure compound 5. Compounds 5d and 5e were then
recrystallized. A similar reaction of 5d (0.89 mmol) with
Selectfluor1 (1.86 mmol) (130 8C, 4 h) gave 6d in 49%
isolated yield. Compound data given in Tables 1 and 2.
In conclusion, we have described the synthesis and
characterization of a series of 3,5-diaryl-4-fluoroisoxazoles
via electrophilic fluorination with Selectfluor1. A unique
4,4,5-trifluoroisoxazoline was also isolated as a side-product
from some of our reactions. Sulfolane was found to be a
good solvent for conducting electrophilic fluorinations with
N-F reagents at higher than normal temperatures. Finally,
this appears to be the first reported example of the direct
fluorination of an isoxazole.
4.3. Fluorination of 4-chloro-3,5-diarylisoxazole 7
A mixture of compound 7 (0.335 g, 1.0 mmol), Select-
fluor1 (0.74 g, 2.1 mmol) and sulfolane (2.0 g) was heated
under nitrogen at 130 8C for 3 h. Dilution with excess water
gave a yellow oil that was separated by decanting. The oil
was dissolved in ether, washed with brine, and dried
(Na2SO4). GC–MS of the extract showed that compounds
8a, 8b, 6d, and 7 were major components (not quantified). A
number of trace products were also present, of which only
5d could be identified. Column chromatography eluting with
pentane gave an inseparable mixture of 8a, 8b, and 6d as a
colorless oil (0.09 g), which solidified after standing several
weeks. 19F NMR (CDCl3): 8a or 8b: ꢁ137.72 (d, J = 4.8),
ꢁ110.63 (d, J = 4.8); 8a or 8b: ꢁ107.94 (d, J = 2.1),
ꢁ101.30 (d, J = 2.7); signals for 6d were also present. GC–
MS data further supported the structure assignments.
4. Experimental
Starting isoxazoles were prepared from the appropriate
chalcone derivative via standard chemistry [21] and
characterized by 13C NMR as outlined in the literature
[22]. Selectfluor1 was either purchased from Aldrich or was
obtained as a gift from Air Products and Chemicals Inc.
Acetonitrile was distilled from calcium hydride. Sulfolane
was dried by standing over activated molecular sieves (4A).
Selecto silica gel (230–400 mesh) was used for column
chromatography. With the exception of the 13C NMR of 6d,
which was obtained on a Bruker 600 MHz instrument, all
NMR spectra were recorded on a Varian Unity+300
instrument at ambient temperature. Chemical shifts for
19F NMR were measured from hexafluorobenzene as
internal standard and converted to the d-scale with CFCl3
as reference by the conversion relation d(CFCl3) = d(C6F6)
ꢁ 162.9 ppm [23]; spectra exhibiting 19F-19F coupling were
run with acquisition time = 1.0 s to improve resolution. GC–
MS data were obtained on a Shimadzu GC-17A/QP-5000
instrument operating in EI mode (70 eV). Other details are
as previously described [6].
Acknowledgments
We thank the Georgia State University Research Fund
and the Georgia Research Alliance for financial support. We
also thank Air Products and Chemicals Inc., for the generous
gift of Selectfluor1 and Dr. Markus Germann for assistance
in obtaining the high-field 13C NMR of compound 6d.
4.1. General procedure for synthesis of 4-fluoro-3,5-
diarylisoxazoles 5a–c
References
A mixture of the isoxazole 4 (1.0 mmol) and Selectfluor1
(1.05 mmol) in dry acetonitrile (5 ml) was heated at gentle
reflux overnight under nitrogen. Excess water was then
added to give a white precipitate, which was filtered, washed
with water and air dried. Compounds 5a and 5b were
purified by preparative plate chromatography (SiO2, 0.5 mm
thickness) eluting with hexanes:ethyl acetate (5:1), while
compound 5c was purified by column chromatography
eluting with 2% ether in pentane. Each compound was then
recrystallized. Compound data given in Tables 1 and 2.
[1] V.S. Parmar, N.K. Sharma, M. Husain, A.C. Watterson, J. Kumar, L.A.
Samuelson, A.L. Cholli, A.K. Prasad, A. Kumar, S. Malhotra, N.
Kumar, A. Jha, A. Singh, I. Singh, A. Himanshu Vats, N.A. Shakil, S.
Trikha, S. Mukherjee, S.K. Sharma, S.K. Singh, A. Kumar, H.N. Jha,
C.E. Olsen, C.P. Stove, M.E. Bracke, M.M. Mareel, Bioorg. Med.
Chem. 11 (2003) 913–929.
[2] B.P. Kansagra, H.H. Bhatt, A.R. Parikh, Indian J. Heterocyclic Chem.
12 (2002) 61–64.
[3] R.E. Gershman, E.R. DeSombre, R.N. Hanson, A. Hughes, in: Pro-
ceedings of the 224th ACS National Meeting, Boston, MA, 18–22
August 2002 (MEDI-385) (Abstracts).
[4] (a) H. Gallardo, C. Zucca, L. Da Silva, Mol. Cryst. Liq. Cryst. 373
(2002) 181;
4.2. General procedure for synthesis of 4-fluoro-3,5-
diarylisoxazoles 5d–e and 4,4,5-trifluoro-3,5-
diarylisoxazolines 6d–e
(b) T. Haino, M. Tanaka, K. Ideta, K. Kubo, A. Mori, Y. Fukazawa,
Tetrahedron Lett. 45 (2004) 2277–2279.
[5] A. Khatyr, H. Maas, G. Calzaferri, J. Org. Chem. 67 (2002) 6705–
6710.
A mixture of the isoxazole 4 (3.0 mmol), Selectfluor1
(3.15 mmol) and sulfolane (7.0 g) was heated under nitrogen
[6] R.A. Day, J.A. Blake, C.E. Stephens, Synthesis (2003) 1586–1590.
[7] F.M.D. Ismail, J. Fluorine Chem. 118 (2002) 27–33.

C.E. Stephens, J.A. Blake / Journal of Fluorine Chemistry 125 (2004) 1939–1945
1945
[8] (a) B.E. Smart, J. Fluorine Chem. 109 (2001) 3–11;
(b) F.C. Krebs, H. Spanggaard, J. Org. Chem. 67 (2002) 7185–7192.
[9] C.L. Bumgardner, J.C. Sloop, J. Fluorine Chem. 56 (1992) 141–146.
[10] (a) G.S. Lal, G.P. Pez, R.G. Syvret, Chem. Rev. 96 (1996) 1737–1755;
(b) S.D. Taylor, C.C. Kotoris, G. Hum, Tetrahedron 55 (1999) 12431–
12477;
[16] P.M. O’Neill, M.D. Tingle, R. Mahmud, R.C. Storr, S.A. Ward, B.K.
Park, Bioorg. Med. Chem. Lett. 5 (1995) 2309–2314.
[17] T. Fujii, S. Asai, T. Okada, W. Hao, H. Morita, T. Yoshimura,
Tetrahedron Lett. 44 (2003) 6203–6205.
[18] J.S. Moilliet, Industrial routes to ring-fluorinated aromatic com-
pounds, in: R.E. Banks, B.E. Smart, J.C. Tatlow (Eds.), Organofluorine
Chemistry, Principles and Commercial Applications, Plenum Press,
New York, 1994, pp. 195–219.
(c) T. Umemoto, S. Fukami, G. Tomizawa, K. Harasawa, K. Kawada,
K. Tomita, J. Am. Chem. Soc. 112 (1990) 8563–8575;
(d) J.J. Hart, R.G. Syvret, J. Fluorine Chem. 100 (1999) 157–161;
(e) K. Adachi, Y. Ohira, G. Tomizawa, S. Ishihara, S. Oishi, J.
Fluorine Chem. 120 (2003) 173–183;
[19] (a) M. Albert, K. Dax, J. Ortner, Tetrahedron 54 (1998) 4839–4848;
(b) S.P. Vincent, M.D. Burkart, C.-Y. Tsai, Z. Zhang, C.-H. Wong, J.
Org. Chem. 64 (1999) 5264–5279.
(f) G.S. Lal, J. Org. Chem. 58 (1993) 2791–2796.
[20] S.D. Sokolov, N.K. Kochetkov, Dokl. Akad. Nauk SSSR 156 (1964)
1391 (also see [14]; p. 338).
[11] M. Zupan, J. Iskra, S. Stavber, J. Fluorine Chem. 70 (1995) 7–8.
[12] M. Zupan, J. Iskra, S. Stavber, Tetrahedron 52 (1996) 11341–11348.
[13] K. Lange, M. Hysell, M. Druelinger, in: Proceedings of the 213th ACS
National Meeting, San Francisco, CA, 13–17 April 1997 (CHED-562)
(Abstracts).
[21] A.L. Baumstark, D.R. Chrisope, R.A. Keel, D.W. Boykin, J. Hetero-
cyclic Chem. 17 (1980) 1719–1721.
[22] D.R. Chrisope, R.A. Keel, A.L. Baumstark, D.W. Boykin, J. Hetero-
cyclic Chem. 18 (1981) 795–798.
[14] P. Grunanger, V.-F. Paola, Isoxazoles Part 1, Wiley, New York, 1991.
[15] R.E. Banks, M.K. Besheesh, R.W. Gorski, N.J. Lawrence, A.J. Taylor,
J. Fluorine Chem. 96 (1999) 129–133.
[23] S.K. Kotovskaya, S.A. Romanova, V.N. Charushin, M.I. Kodess, O.N.
Chupakhin, J. Fluorine Chem. 125 (2004) 421–428.