Photochemical Schiemann reaction in ionic liquids

Article abstract of DOI:10.1016/j.jfluchem.2007.03.012

Photochemical Schiemann reactions of imidazole derivatives 1 and 4 were carried out in 1-butyl-3-methylimidazolium tetrafluoroborate ionic liquid [bmim][BF₄] as solvent. The effects of temperature, co-solvent and wavelength on the rate of the reaction and product yield were examined. The use of ionic liquid increases the yield of the photochemical fluorodediazoniation reaction of 2 at 0 °C. Careful temperature control is necessary to minimize the photodecomposition of the ionic liquid in order to increase the yield of product.

Full text of DOI:10.1016/j.jfluchem.2007.03.012

Journal of Fluorine Chemistry 128 (2007) 674–678
www.elsevier.com/locate/fluor
Short communication
Photochemical Schiemann reaction in ionic liquids
*
Jorge Heredia-Moya, Kenneth L. Kirk
Laboratory of Bioorganic Chemistry, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health,
Department of Health and Human Services, Bethesda, MD 20892, United States
Received 29 January 2007; received in revised form 9 March 2007; accepted 10 March 2007
Available online 16 March 2007
Abstract
Photochemical Schiemann reactions of imidazole derivatives 1 and 4 were carried out in 1-butyl-3-methylimidazolium tetrafluoroborate ionic
liquid [bmim][BF4] as solvent. The effects of temperature, co-solvent and wavelength on the rate of the reaction and product yield were examined.
The use of ionic liquid increases the yield of the photochemical fluorodediazoniation reaction of 2 at 0 8C. Careful temperature control is necessary
to minimize the photodecomposition of the ionic liquid in order to increase the yield of product.
# 2007 Elsevier B.V. All rights reserved.
Keywords: Photochemistry; Ionic liquid; Imidazole; Schiemann reaction
1. Introduction
and fluoroimidazoles were delivered in satisfactory yields [2].
Advantages of this procedure include its generality and the fact
that it can be done in situ without isolation of the diazonium salt
[3]. Using this photochemical variation of the Schiemann
reaction, we subsequently prepared many fluorinated analogues
of biologically important imidazoles for extensive biological
evaluation [4].
The Balz–Schiemann reaction was one of the first methods
developed for the synthesis of fluorinated aromatic compounds
and is still widely used. In this reaction the fluorinated
compound is formed by thermal decomposition of an aryl
diazonium tetrafluoroborate or hexafluorophosphate salt [1].
This is a valuable transformation that has made available many
new aryl fluorides. However, the original procedure had
reproducibility problems and gave yields that were quite
dependent on the arene substrate. Furthermore, the high
temperatures often required to effect cleavage of the diazonium
salt carbon–nitrogen bond can lead to thermal destruction of
product or starting material. Thus, in repeated attempts to
prepare ring-fluorinated imidazoles by thermal decomposition
of imidazole diazonium fluoroborates, we were unable to detect
any trace of desired fluorinated product despite destruction of
starting material [2].
A limitation of the photochemical Schiemann reaction stems
from competing reaction with solvent nucleophiles. Studies of
the dediazoniation reactions of diazonium salts have revealed a
heterolytic mechanism for both photochemical and thermal
dediazoniation [5]. Thus, the fate of the aryl cation formed in
the reaction can be strongly affected by the solvent. As an
example, water and alcohols can compete with fluoride as a
nucleophile and form phenols and aryl ethers, respectively
[5a,6].
Ionic liquids are suitable media for stabilization of charged
transition states favoring those mechanisms occurring through
charge separation [7]. We reasoned that in the dediazoniation
reaction the absence of protic nucleophiles in ionic liquids
should eliminate side reactions that result from competing
reactions with solvent and thus improve the yield of the
fluoroproduct.
Faced with lack of success with thermal decomposition of
imidazole diazonium fluoroborates we explored the option of
providing selective energy to the diazonium chromophore by
carrying out a photochemical reaction. We found that under
very mild conditions, UV irradiation promoted efficient
extrusion of nitrogen, thermal side reactions were avoided,
A few studies have recently addressed the use of ionic
liquids in photochemistry, focusing on use of photochemical
probes such as Nile Red to assess polarity [8] in studies of
photoinduced electron transfer features [7,9] in characteriza-
tion of radical ions [10] and in photochemical degradation of
* Corresponding author.
E-mail address: kennethk@intra.niddk.nih.gov (K.L. Kirk).
0022-1139/$ – see front matter # 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.jfluchem.2007.03.012

J. Heredia-Moya, K.L. Kirk / Journal of Fluorine Chemistry 128 (2007) 674–678
675
environmental contaminants [11]. However, their use in
preparative organic photochemistry has been limited to
dimerization [12], photoisomerization [13], and photoreduction
reactions [14].
In the literature, imidazolium ionic liquids are generally
regarded as optically transparent liquids possessing no
absorption in the UV–vis range (above 250 or 300 nm).
Any absorption beyond this wavelength has been ascribed to
the presence of impurities. However, recent studies not only
suggest that these ionic liquids have significant absorption in
the UV region, but also that they have fluorescence that
covers a large part of the visible region [15]. The lack of
transparency of most imidazolium ionic liquids below 320 nm
may explain the limited use of ionic liquids in photochemistry
[9a].
Scheme 1. One-pot diazotization-photochemical fluorodediazoniation in
[bmim][BF4].
With this procedure, the fluoroimidazole 3 was reported in 39%
yield [2a]. The results of reaction in ionic liquid (Scheme 1) are
described below.
2.1. Effect of temperature
In this paper, we report our studies on the photochemical
fluorodediazoniation of imidazole derivatives in ionic liquid.
We have examined the effects of temperature and wavelength
on the rate of the reaction and product yield.
In our first attempt (Scheme 1), we reduced the amount
of water in the media by replacing aqueous sodium nitrite
with n-BuONO in an equivalent of 48% HBF4 and used
[bmim][BF4] > 97% of purity to replace 48% aqueous HBF4
(Table 1, entries 1–5) as solvent. This reaction initially was
done at room temperature in a micro photochemical reactor
using a quartz immersion well. After 1.5 h the initial yellow
color of the solution changed to red and the yield of 3 was
only of 11%. The same yield was obtained using an
equivalent of 48% HBF4 and 1.5 equivalents of n-BuONO.
When the temperature of the reaction was reduced to À78 8C
the mixture solidified and the evolution of nitrogen was
almost negligible. Every hour the mixture was warmed 0 8C
to allow mixing and removal of generated nitrogen and then
cooled again to À78 8C for continued irradiation. After 9 h of
reaction the solution remained yellow and 3 was obtained in
41% yield. The reaction at 0 8C affords 3 in 30% yield after
1 h of reaction.
2. Results
Previous attempts to carry out the thermal Balz–Schiemann
reaction of imidazole-4-diazoniumfluoroborate-5-carboxylic
esters 2 resulted in decomposition of starting material but no
trace of fluorinated imidazole could be detected [2]. However,
photochemical decomposition proceeded smoothly and the
photochemical decomposition of 2 provided the first synthesis
of ring fluorinated imidazoles. This reaction remains the only
general route to this important class of compounds. The initial
procedure consisted of reaction of a solution of the amine 1 in
HBF4 with NaNO2 at 0 8C followed by irradiation of the
mixture with a UV lamp until evolution of nitrogen ceases.
Table 1
Diazotization-fluorodediazoniation of 2 in [bmim][BF4] (Quartz jacketed immersion well, unless indicated)
Entry
Amine
Solvent
Nitrile source
T (8C)
Time (h)
Yielda (%)
HBF4/n-BuONO
NOBF4
1
2
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
4
[bmim][BF4]
1.2/1.20
–
rt
1.5
3
11
11
30
20
41
31
37
52
51
57
58
62
45
44
46
70
63
56c
[bmim][BF4]
1.0/1.5
–
rt
3
[bmim][BF4]
1.5/1.5
–
0
1
4
[bmim][BF4]
1.5/1.5
–
À10
À78
0
1
5
[bmim][BF4]
1.5/1.5
–
9
6
[bmim][BF4]
–
–
–
–
–
–
–
–
–
–
–
–
–
1.1
1.1
1.5
1.7
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
1.5
9
7
[bmim][BF4]
0
11
2
8
[bmim][BF4]
0
9
[bmim][BF4]
0
6
10
11
12
13
14
15
16
17
18
[bmim][BF4]
0
8
[bmim][BF4]
0
9
[bmim][BF4]
0
24
6
Acetonitrile:[bmim][BF4] 1:1
Acetonitrile:[bmim][BF4] 1:1
Acetonitrile:[bmim][BF4] 1:1
[bmim][BF4]b
0
0
8
À50
0
6
168
168
24
[bmim][BF4]b
rt
[bmim][BF4]
0
a
Yield of 3.
Borosilicate glass jacketed immersion well.
b
c
Yield of 5.

676
J. Heredia-Moya, K.L. Kirk / Journal of Fluorine Chemistry 128 (2007) 674–678
2.2. Source of nitrite
In a similar study of the thermal Schiemann reaction, Laali
and Gettwert [16] used NOBF4 in the diazotization reaction.
This permits complete elimination of water in the reaction
mixture as a strategy to increase the yield, a strategy we adopted
in our next experiments. We explored both the thermal and the
photochemical reactions.
Scheme 2. Diazotization-photochemical fluorodediazoniation reaction of 4 in
[bmim][BF4].
In the thermal reaction of 2 using 1.5 equiv. of NOBF4 at
110 8C no reaction was observed after 5 h. The tlc showed only
starting material. After 12 h of reaction at 120 8C no reaction
was observed, but the ionic liquid had visibly decomposed. The
thermal stability of imidazolium based ionic liquids will be
significantly reduced in the presence of nucleophiles, especially
at extended times at high temperature [17] or in the presence of
fluoride ion [18]. After 7 h at 140 8C the diazonium salt
disappeared but no fluoroimidazole 3 was observed. This result
is consistent with our original observations during attempts to
synthesize ring-fluorinated imidazoles by a thermal Schiemann
reaction [2].
isolated as imidazole cyclic trimer 6 that corresponded to a
yield of 5 of 56% (Scheme 2). In an attempt to improve the
solubility of the sulfate in [bmim][BF4], a small amount of
water was added, but this resulted in a lower yield.
3. Discussion
The use of ionic liquids in the photochemical fluorodedia-
zoniation of diazonium salt 2 has been examined as a strategy to
increase the yield of fluoroimidazole 3 by eliminating side
reactions related to competing reactions with protic solvents.
We found that temperature control is critical and that the
reaction must be carried out below 0 8C if a quartz well is used,
otherwise, photodecomposition of the ionic liquid increases
and the yield of product decreases. The photodecomposition of
the ionic liquid can be reduced using a borosilicate well but this
necessitates an increase in reaction time. This increase in
reaction time does not affect the yield.
The ionic liquid used is a mobile liquid at room temperature,
but at 0 8C it becomes very viscous. However, it remains mobile
enough to permit stirring and evolution of nitrogen during the
irradiation. Using 1.5 equiv. of NOBF4, the reaction affords 3 in
62% yield after 1 day (Table 1, entry 12). When the equivalents
of NOBF4 are less than 1.5 (Table 1, entries 6 and 7) the yield
decreased and when higher than this (Table 1, entry 9) the yield
did not change, even with a higher time of reaction.
One of the major advantages of ionic liquids is that they can
be easily recycled. However, recent studies have shown that
imidazolium-based ionic liquids suffer degradation by UV-light
and the degree of degradability is dependent on the nature of n-
alkyl chain [20]. This degradation increases with the duration of
irradiation [11a] and is enhanced by the presence of reactive
species [20].
2.3. Addition of co-solvent
In order to decrease viscosity, we used a mixture of
acetonitrile: [bmim][BF4] 1:1 as solvent (Table 1, entries 13–
15). In this case the yields of 3 after 6 h at 0 8C and À50 8C
were 45% and 46%, respectively. Additionally, at À50 8C it was
possible to obtain 1% yield of a solid that was identified as 5-
The presence of BF3 and HF in the reaction could produce
new species similar to those observed in N-methylimidazole
treated with HCl or CF3COOH [15] that could absorb incident
photons. These new species could interfere in the photo-
chemical reaction and favor the photodecomposition of the
ionic liquid especially with temperatures over 0 8C. Even at
0 8C the absorbance of [bmim][BF4] before and after the
1
diazoimidazole-4-carboxylic acid ethyl ester by H NMR and
MS.
2.4. Wavelength
We also investigated the effect of wavelength on the
reaction (Table 1, entries 16 and 17). Using a borosilicate glass
jacketed immersion well in the photochemical reactor led to a
slower rate of conversion but increased the yield. After 7 days
of reaction at 0 8C, 3 was obtained in a 70% yield. In addition,
under these conditions of longer wavelength irradiation it was
possible to carry out the reaction at room temperaturewith only
a small decrease in the yield (63%) after the same reaction
time.
2.5. Alternate substrate
A suspension of 2-aminoimidazole sulfate (4) was treated
with the NOBF4 and irradiated at 0 8C overnight using a quartz
jacketed immersion well to afford 2-fluoroimidazole 5 (Table 1,
entry 18). Due the reactivity of 2-fluoroimidazole [19], it was
Fig. 1. Absorbance of [bmim][BF4]: (a) before and (b) after irradiation.

J. Heredia-Moya, K.L. Kirk / Journal of Fluorine Chemistry 128 (2007) 674–678
677
reaction (Fig. 1) show the same pattern observed in the
5.3. Imidazole Cyclic Trimer (6)
photodecomposition of 1-ethyl-3-methylimidazolium bis(per-
fluoroethylsulfonyl) imide ([emim][beti]) and [bmim][PF6]
[11a].
To a solution of 308.9 mg of 2-aminoimidazole sulfate 98%
(4) (194.3 mg, 2.29 mmol) in 2.0 mL of [bmim][BF4] at 0 8C
was added slowly a cooled suspension of 399.9 mg of NOBF4
(3.42 mmol) in 1.0 mL of [bmim][BF4] while N2 was bubbled
into the solution. Additional [bmim][BF4] (1.0 mL) was used to
transfer all the NOBF4 into the reactor. After 90 min, the
reaction mixture was irradiated overnight using the quartz
jacketed immersion well. Water (10 mL) was added and the
solution was extracted with ether. The organic layers were dried
over Na2SO4 and the solvent was removed under vacuum. The
residue was heated to complete conversion of 2-fluoro-1H-
imidazole (5) to trimer 6 and then purified by column
chromatography (KP-Sil column, 20–50% MeOH in CH2Cl2,
UV 220–245 nm) to afford 85.1 mg of imidazole cyclic trimer 6
that corresponds to 110.9 mg, 56% of 2-fluoro-1H-imidazole
(5). The 1H NMR and MS were in complete agreement with the
data reported in literature [19].
The presence of impurities in the ionic liquids clearly could
affect the rate and product formation in our photochemical
procedure, for instance by interfering with light transmission in
the solution. It is interesting that competition between
impurities and ionic liquid for photon absorption has been
cited as a protective factor that enhances the stability of the
solvent during irradiation [11a]. We have not investigated these
factors in our own work.
4. Conclusion
The photochemical Schiemann reaction in ionic liquid leads
to a somewhat improved yield compared to the reaction done in
aqueous HBF4. On the other hand, technical problems related to
viscosity at low temperatures suggest that the increased yield
may not outweigh other disadvantages.
Acknowledgement
5. Experimental
This work was supported by the intramural research program
of NIDDK, NIH.
5.1. General
All the reagents were from Aldrich and used without further
purification. NMR spectra were run in CDCl3 on a Varian
Gemini 300 MHz spectrometer. Mass spectra were determined
using a Jeol SX-102 instrument. The temperature was
controlled with a Neslab CC100 cooling machine. All the
reaction were made in a micro photochemical reactor using a
jacketed immersion well, in either quartz or borosilicate glass
and a quartz Pen-Ray 5.5 W, low pressure, cold cathode,
mercury lamp.
References
[1] G. Balz, G. Schiemann, Chem. Ber. 60 (1927) 1186.
[2] (a) K.L. Kirk, L.A. Cohen, J. Am. Chem. Soc. 93 (1971) 3060–3061;
(b) K.L. Kirk, L.A. Cohen, J. Am. Chem. Soc. 95 (1973) 4619–4624;
(c) K.L. Kirk, W. Nagai, L.A. Cohen, J. Am. Chem. Soc. 95 (1973) 8389–
8392.
[3] B. Langlois, in: B. Baasner, H. Hagemann, J.C. Tatlow (Eds.), Methods of
Organic Chemistry (Houben-Weyl). Organo-Fluorine Compounds, vol.
10a, Thieme, Stuttgart, 2000, pp. 686–740.
[4] K.L. Kirk, L.A. Cohen, in: R. Filler (Ed.), Biochemistry of the Carbon-
Fluorine Bond, ACS Symposium Series, 1976, 23–36.
5.2. 5-Fluoro-1H-imidazole-4-carboxylic acid ethyl ester
(3)
[5] (a) A. Milanesi, M. Farnoni, A. Albini, J. Org. Chem. 70 (2005) 603–610;
(b) P.S.J. Canning, K. McCrudden, H. Maskill, B. Sexton, J. Chem. Soc.,
Perkin Trans. 2 (1999) 2735–2740.
[6] K.L. Kirk, J. Org. Chem. 41 (1976) 2373–2376.
´
[7] M. Alvaro, E. Carbonell, B. Ferrer, H. Garcıa, J. Herance, J. Photochem.
To a solution of 226.8 mg of 5-amino-1H-imidazole-4-
carboxylic acid ethyl ester (1) (1.46 mmol) in 2.0 mL of
[bmim][BF4] at 0 8C was added slowly a cooled suspension of
255.2 mg of NOBF4 (2.18 mmol) in 1.0 mL of [bmim][BF4]
while N2 was bubbled into the solution. Additional
[bmim][BF4] (1.0 mL) was used to transfer all the NOBF4
into the reactor. After 90 min, the reaction mixture was
irradiated using the quartz (overnight) or borosilicate glass (7
days) jacketed immersion well. Water (10 mL) was added and
the solution was extracted with ether. The organic layers were
dried over Na2SO4 and the solvent was removed under vacuum.
The residue was purified by column chromatography (KP-Sil
column, 20–50% EtOAc in Hexane, UV 235–245 nm) to afford
142.9 mg, 62% (quartz) and 153.9 mg, 70% (borosilicate glass)
of 5-fluoro-1H-imidazole-4-carboxylic acid ethyl ester (3) as a
white solid. The 1H NMR and MS were in complete agreement
with the data reported in the literature [2a]. Traces of 5-
diazoimidazole-4-carboxylic acid ethyl ester also were found
(see results).
Photobiol. 82 (2006) 185–190.
[8] A.J. Carmichael, K.R. Seddon, J. Phys. Org. Chem. 13 (2000) 591–595.
´
[9] (a) S. Marquis, B. Ferrer, M. Alvaro, H. Garcıa, H. Roth, J. Phys. Chem. B
110 (2006) 14956–14960;
´
(b) M. Alvaro, B. Ferrer, H. Garcıa, M. Narayana, Chem. Phys. Lett. 362
(2002) 435–440;
(c) C.M. Gordon, A.J. McLean, Chem. Commun. (2000) 1395–1396;
(d) C. Lee, T. Winston, A. Unni, R.M. Pagni, G. Mamantov, J. Am. Chem.
Soc. 118 (1996) 4919–4924.
[10] A. Marcinek, J. Zielonka, J. Ge˛bicki, C.M. Gordon, I.R. Dunkin, J. Phys.
Chem. A 105 (2001) 9305–9309.
[11] (a) Q. Yang, D.D. Dionysiou, J. Photochem. Photobiol. A: Chem. 165
(2004) 229–240;
(b) Q. Yang, D.D. Dionysiou, in: R.D. Rogers, K.R. Seddon (Eds.), Ionic
Liquids IIIB: Fundamentals, Progress, Challenges, and Opportunities,
ACS Symposium Series 902, American Chemical Society, Washington,
DC, 2005, pp. 182–198.
[12] G. Hondrogiannis, C.W. Lee, R.M. Pagni, G. Mamantov, J. Am. Chem.
Soc. 113 (1993) 9828–9829.
[13] (a) S.C. Hubbard, P.B. Jones, Tetrahedron 61 (2005) 7425–7430;
(b) J. Ding, V. Desikan, X. Han, T.L. Xiao, R. Ding, W.S. Jenks, D.W.

678
J. Heredia-Moya, K.L. Kirk / Journal of Fluorine Chemistry 128 (2007) 674–678
Armstrong, Org. Lett. 7 (2005) 335–337;
(c) R. Ozawa, H. Hamaguchi, Chem. Lett. 30 (2001), 736-736.
[18] C.B. Murray, G. Sandford, S.R. Korn, J. Fluorine Chem. 123 (2003) 81–
84.
[14] J.L. Reynolds, K.R. Erdner, P.B. Jones, Org. Lett. 4 (2002) 917–919.
[15] A. Paul, P.K. Mandal, A. Samanta, Chem. Phys. Lett. 402 (2005) 375–379.
[16] K.K. Laali, V.J. Gettwert, J. Fluorine Chem. 107 (2001) 31–34.
[17] A.G. Glenn, P.B. Jones, Tetrahedron Lett. 45 (2004) 6967–6969.
[19] Y. Takeuchi, K.L. Kirk, L.A. Cohen, J. Org. Chem. 44 (1979) 4243–
4246.
[20] P. Stepnowski, A. Zaleska, J. Photochem. Photobiol. A: Chem. 170 (2005)
45–50.