Improved synthesis of d,l-fluorocitric acid

Article abstract of DOI:10.1016/j.jfluchem.2011.07.006

We propose a decagram synthesis of commercially unavailable fluorocitric acid. The synthesis begins with benzyl fluoroacetate, which is converted to dibenzyl 2-fluoro-3-oxosuccinate, followed by condensation with malonic acid or its monobenzyl ester and subsequent hydrogenolysis.

Full text of DOI:10.1016/j.jfluchem.2011.07.006

Journal of Fluorine Chemistry 132 (2011) 920–924
Contents lists available at ScienceDirect
Journal of Fluorine Chemistry
journal homepage: www.elsevier.com/locate/fluor
Improved synthesis of
D
,L-fluorocitric acid
*
Ruben Vardanyan , Vlad K. Kumirov, Victor J. Hruby
Department of Chemistry and Biochemistry, University of Arizona, Tucson, AZ 85721, USA
A R T I C L E I N F O
A B S T R A C T
Article history:
We propose a decagram synthesis of commercially unavailable fluorocitric acid. The synthesis begins
with benzyl fluoroacetate, which is converted to dibenzyl 2-fluoro-3-oxosuccinate, followed by
condensation with malonic acid or its monobenzyl ester and subsequent hydrogenolysis.
ß 2011 Published by Elsevier B.V.
Received 4 May 2011
Received in revised form 8 July 2011
Accepted 11 July 2011
Available online 20 July 2011
Keyword:
Fluorocitric acid
1. Introduction
with ethyl fluoroacetate to give oxalylfluoroacetic ester 1 and, as a
minor product, a,
a0-difluorocitric ethyl ester 2 [19]. The previously
Traditionally, clinicians have treated unresponsive patients by
prescribing several different drugs to activate multiple therapeutic
mechanisms. An alternative strategy would be implementation of
a single multifunctional chemical entity that is able to modulate
multiple targets simultaneously. The basic principles and meth-
odologies for the design of multifunctional drugs have been
reviewed [1–6].
Evidence has been provided that the activation of astrocytes in
the spinal cord is involved in the modulation of neuropathic pain.
An inhibitor of glia metabolism – barium fluorocitrate reduces
astrocyte activation and significantly attenuated the development
of pain hypersensitivity [7–18].
We hypothesized that bifunctional investigational tools could be
representedasioniccompounds–pharmaceuticalsalts, whichcould
be composed by an interaction of any opioid agonist as the organic
base with the classical glial inhibitor, fluorocitric acid, as the acid.
A possible advantage of this approach would be that after
dissociation, both counterions will be able to act simultaneously
and independently at their appropriate targets. The main barrier
for checking this hypothesis is the commercial unavailability of
fluorocitric acid. Here we describe the design and the methodology
of a large scale (multiple grams) method for the preparative
synthesis of fluorocitric acid.
described procedure was modified to give high yields of pure
oxalylfluoroacetic ester 1, which then was used as a carbonyl
component in the Reformatsky reaction with ethyl zincbromoace-
tate. The Reformatsky reaction led to a multicomponent mixture,
from which a small amount of pure product – triethyl fluorocitrate 3
was separated and hydrolyzed under acidic conditions to give
fluorocitric acid. In contrast to published studies, the product could
beisolated only asitsbarium salt. Analogous resultswerepreviously
described by hydrolysis under basic conditions using 100% excess of
sodium hydroxide, then passing the basic solution through a strong
cation-exchange column [20]. In addition to the problem of low
yields of triethyl fluorocitrate 3 in the Reformatsky reaction, there
are two additional problems when utilizing this method. The first
problem concerns transformations 3 ! 4 and 3 ! 5 (Scheme 1), in
whichitisveryhardtoachievethefullhydrolysisofthetriproticacid
ester. We have tried to implement a variety of methods and
conditions of acidic and basic hydrolysis, changing acids, bases,
solvents and temperatures, but could not obtain pure fluorocitric
acid 5 in satisfactory yields. The second problem concerns the very
high hydrophilicity of obtained product, which makes it practically
impossible to perform an efficient extraction intoan organicsolvent.
Another approach for the synthesis of fluorocitric acid 5 was
undertook implementing a method of synthesis [21,22], in which
instead of a Reformatsky reaction, the Doebner modification of the
Knoevenagel reaction of oxalylfluoroacetic ester 1 with malonic
acid is employed (Scheme 2). We have extended this reaction, and
for the synthesis of the triethyl ester of fluorocitric acid 3, instead
of malonic acid, its monoethyl ester was used. This approach
allowed us to obtain excellent yields of pure 3 and 6, but hydrolysis
of both triethyl ester 3 and diethyl ester 6 faced the same problems
and we could isolate the fluorocitric acid 5 only as a barium salt
with low yields.
2. Results and discussion
We started with the first published method of the synthesis of
fluorocitricacid (Scheme 1), in which diethyl oxalatewas condensed
* Corresponding author.
E-mail address: vardanyr@email.arizona.edu (R. Vardanyan).
0022-1139/$ – see front matter ß 2011 Published by Elsevier B.V.
doi:10.1016/j.jfluchem.2011.07.006

R. Vardanyan et al. / Journal of Fluorine Chemistry 132 (2011) 920–924
921
Scheme 1. Implementation of the Reformatsky reaction in the synthesis of fluorocitric acid.
However, we found that replacing the hydrolysis procedure by
hydrogenation could be a good solution to the problem (Scheme 3),
andforthispurposethesynthesisofdibenzyl8andtribenzyl9esters
of fluorocitric acid were developed. The starting material for their
synthesis – dibenzyl 2-fluoro-3-oxosuccinate 7 has been prepared
from dibenzyloxalate and benzyl fluoroacetate (Scheme 4).
Scheme 4. Synthesis of dibenzyloxalylfluoroacetic ester.
Condensation of dibenzyl 2-fluoro-3-oxosuccinate
7
with
malonic acid and the malonic acid monobenzyl ester in pyridine
gave good yields of the dibenzyl 8 and tribenzyl 9 esters of
fluorocitric acid. Hydrogenolysis of both of compounds gave
excellent yields of fluorocitric acid 5.
Compounds 2, 3, 5, 6, 7, 8, 9 had interesting 1H and 13C NMR
spectra due to spin–spin coupling of 19F, a spin- nucleus. The
observed 1H and 13C signals gave doublets by 19F, through 1–4
bonds. The following coupling constants were observed:
2
3
2
4
1JCF ꢀ 200 Hz, JCF ꢀ 20 Hz, JCF ꢀ 5 Hz, JHF ꢀ 50 Hz, JHF ꢀ 0–
2
3
5 Hz, JHH ꢀ 10–15 Hz, JHH ꢀ 6–8 Hz. The magnitude of the
coupling constant due to fluorine increases with proximity to
the 19F nucleus, which aided in the assignment of carbonyl
resonances in the 13C NMR.
Scheme 2. Implementation of the Knoevenagel–Doebner reaction in the synthesis
Compound 5 was characterized by 1H, 13C, and 19F NMR and by
1H–13C HSQC. In the 1H spectrum, the OH protons were not
observed due to exchange with CD3OD and one of the methylene
protons was observed to have a 4JHF coupling of 0.9 Hz. In the 1H-
decoupled HSQC spectrum, the 1H–13C(19F) showed two cross-
peaks in an anti-diagonal correlation. The separation between the
of fluorocitric acid.
2
peaks in the F2 dimension corresponds to the JHF coupling
(47.4 Hz) and the separation between the peaks in the F1
1
dimension corresponds to the JCF coupling (ꢁ193.2 Hz). Both
coupling constants are consistent with the magnitude and sign of
known constants [23]. 19F NMR spectrum showed five doublets of
2
different intensities (25:1:90:15:15) with JHF = 47.5 Hz.
3. Conclusions
Bulk synthesis of commercially unavailable fluorocitric acid has
been proposed starting from readily synthesized dibenzyl 2-
fluoro-3-oxosuccinate, followed by condensation with malonic
acid or its monobenzyl ester and subsequent hydrogenolysis.
Scheme 3. Synthesis of tribenzyl 9 and dibenzyl 8 esters of fluorocitric acid.

922
R. Vardanyan et al. / Journal of Fluorine Chemistry 132 (2011) 920–924
4. Experimental
4.1. General remarks
(s), 62.33 (s), 13.87 (s), 13.80 (s). EI-MS: m/z 381; HRMS calcd for
23H29N2O3: 381.2178; found (ESI, [M+H]+): 381.2183.
C
4.3. General procedure for Reformatsky reaction: D,L-fluorocitric acid
The compounds were characterized by 1H and 13C NMR. All
standard 1H and 13C NMR experiments were performed on
Bruker DRX-600 equipped with a Narolac 5 mm probe at 298 K.
triethyl ester (3)
4.3.1. Method 1
19F NMR was performed on Varian Unity-300 using 10
m
L
A typical 0.05 mol scale procedure was performed as follows.
The activated Zn metal dust 6.4 g (0.1 mol) was placed in a reaction
flask, its surface was covered with dry THF, a couple of crystals of
iodine, and a couple of drops of ethyl 2-bromoacetate was added
and the mixture was heated up to boiling temperature of THF and
the vapours of iodine disappeared. At that moment a solution of
oxalylfluoroacetic ester 1 10.3 g (0.05 mol) was added in 30 mL of
THF and on stirring and boiling 9.185 g (0.055 mol) of ethyl 2-
bromoacetate in equal volume of THF was added dropwise during
30 min. The reaction initiated typically after addition of 20–30% of
the total amount of ethyl 2-bromoacetate and started to boil very
violently. It is obvious that this process could not be scaled up due
to its unpredictable point of initiation, which made efficient
cooling impossible. The reaction mixture was heated to reflux
additional one hour until obtaining a clear yellowish-orange
solution and left for the night at room temperature. Next day it was
filtered, THF was evaporated and 75 mL of chloroform was added
to remaining residue. The mixture was cooled (ice bath) and on
stirring 0.2 mol of H2SO4 as a 5% solution was added. The
chloroform solution was separated, dried on MgSO4 and the
solvent was evaporated. Attempts to distill the remaining
polycomponent mixture (TLC) resulted in intensive tarification.
Column chromatography also was inefficient. A small quantity 1–
1.5 g (7–10%) of product 3 was crystallized during a couple of days,
which was separated via filtration affording the desired Refor-
matsky products in pure form and used in following steps. Ten
experiments have been performed with changes of solvents, order
of adding of reagents, etc., which confirmed that the proposed
procedure was not practical of for preparative scale preparations.
trifluoroacetic acid as an internal standard referenced at
ꢁ78.5 ppm. Multiple resonances were observed in some spectra,
due to conformers, diastereomers, or different protonation
states; the largest signal was reported for chemical shift and
multiplicity. Chemical shifts for 1H and 13C NMR are reported
with respect to TMS at 0.00 ppm. Coupling constants are
reported in hertz. Fluorine is a spin- nucleus and was not
decoupled for NMR experiments.
The compounds were analyzed by GC–MS measurement using a
Micromass GCT instrument (with a DB5 column) with standard GC
and electron impact (EI) ionization (70 eV) conditions. In most
cases, the molecular ion was not detected, thus electrospray
ionization (ESI) was used to determine accurate masses for the
protonated molecules ([M+H]+) on a Bruker 9.4 T Apex Qh Fourier
transform ion cyclotron resonance (FT-ICR) instrument. The
accuracy of these measurements, i.e., the chemical formulae
determinations, were <2.0 ppm.
The purity of compounds was determined by TLC on silica gel
plates (Analtech 02521), solvent system MeOH:CHCl3 1:4. Melting
points are uncorrected.
4.2. General procedure for the synthesis of oxalylfluoroacetic diethyl
ester (1)
To a cooled (ice bath) stirred solution of sodium ethoxide,
prepared from 3.105 g (0.135 mol) of sodium and 75 mL of dry
ethanol was added dropwise diethyl oxalate 14.6 g (0.1 mol) in
equal volume of dry ethanol and after 15 min, a solution of 10.6 g
(0.1 mol) ethyl fluoroacetate1 in 15 mL of dry ethanol was added.
The mixture was stirred additional 2–3 h allowing come to room
temperature and left for a night. Next day 75 mL of dry ether was
added to the mixture, precipitated salt was filtered off washed with
dry ether and placed in 150 mL of ether. The mixture was cooled
(ice bath) and the obtained salt was hydrolyzed with 10% solution
of HCl (0.135 mol). Ether layer was separated, dried on MgSO4 and
after evaporation of the solvent products were distilled giving two
fractions. Both of the fractions are identical to previously described
samples [1–3].
4.3.2. Method 2
Oxalylfluoroacetic ester 1 (7.725 g, 0.0375 mol) was added over
a period of 30 min to a solution of malonic acid monoethyl ester
(4.95 g, 0.0375 mol) in 10 mL of pyridine. The reaction mixture was
stirred at room temperature for 2 h, then heated on a steam bath
for ꢀ30 min until evolution of CO2 ceased. The pyridine was
evaporated, and 75 mL of ether was added to the remaining
residue. The traces of pyridine was extracted with 5% HCl 2ꢂ
10 mL, excess of malonic acid monoethyl ester washed out with 5%
NaHCO3 solution 2ꢂ 10 mL. This organic phase was dried with
MgSO4. After removal of ether 7.7 g (70%) of pure triethyl
monofluorocitrate 3 was obtained.
4.2.1. Analysis of Fraction 1
Oxalylfluoroacetic diethyl ester 1. 110–112 8C (5 mm Hg) 17.5 g
(69.4%).
1H NMR (600 MHz, CDCl3)
d
5.96 (d, 47.3 Hz, 1H), 4.48–4.36 (m,
2H), 4.36–4.25 (m, 2H), 1.44–1.36 (m, 3H), 1.34–1.27 (m, 3H).
13C NMR (151 MHz, CDCl3)
183.77 (d, 20.6 Hz), 163.08 (d,
4.3.3. Spectroscopic data for
D
,
d
L
-fluorocitric acid triethyl ester (3)
5.10 (d, 47.3 Hz, 1H), 4.32 (q 7.2 Hz,
1H NMR (600 MHz, CDCl3)
d
2H), 4.26 (q, 7.8 Hz, 2H), 4.15 (q, 7.2 Hz, 2H), 3.21–2.85 (m, 2H),
1.33 (t, 7.1 Hz, 3H), 1.30 (t, 7.6 Hz, 3H), 1.25 (t, 7.2 Hz, 3H). 13C NMR
23.5 Hz), 159.49(s), 88.16(d, 197.3 Hz), 63.43(s), 63.14(s), 13.85(s).
(151 MHz, CDCl3)
d 170.73 (d, 6.1 Hz), 169.39 (s), 165.90 (d,
4.2.2. Analysis of Fraction 2
24.7 Hz), 90.90 (d, 197.4 Hz), 76.15 (d, 19.9 Hz), 62.94 (s), 62.03 (s),
61.10 (s), 39.60 (d, 6.2 Hz), 13.97 (s), 13.92 (s), 13.88 (s). [M+H]+,
a,
a0-Difluorocitric ethyl ester 2. 140–160/5, 4.5 g (12.6%).
Crystalline solid, m.p. 68–69 8C.
1H (600 MHz, CDCl3):
5.39 (d, 47.5 Hz, 1H), 5.29 (d, 47.5 Hz,
1H), 4.44–4.22 (m, 6H), 1.41–1.28 (m, 9H). 13C NMR (151 MHz,
CDCl3) 168.55 (d, 6.7 Hz), 167.55 (t, 5.1 Hz), 166.09 (d, 23.4 Hz),
C12H20FO7, calculated: 295.11876; measured: 295.11909 (1.1 ppm
error).
d
d
4.4. Synthesis of D,L-fluorocitric acid diethyl ester (6)
165.82 (d, 20.0 Hz), 165.66 (d, 20.1 Hz), 88.88 (dd, 196.4, 5.0 Hz),
88.45 (dd, 158.3, 3.2 Hz), 87.90 (d, 6.4 Hz), 87.25 (dd, 197.7,
6.4 Hz), 78.64 (t, 19.9 Hz), 77.79 (dd, 22.0, 19.5 Hz), 63.71 (s), 62.43
4.4.1. General procedure for synthesis of D,L-fluorocitric acid diethyl
ester (6)
Oxalylfluoroacetic ester 1 (15.4 g, 0.075 mol) was added over a
period of 30 min to a solution of malonic acid (8.0 g, 0.075 mol) in
1
Caution: Ethyl fluoroacetate is a highly toxic compound.

R. Vardanyan et al. / Journal of Fluorine Chemistry 132 (2011) 920–924
923
20 mL of pyridine. The reaction mixture was stirred at room
temperature for 2 h, then heated on a steam bath for ꢀ30 min until
evolution of CO2 ceased. Pyridine was evaporated, and 75 mL of
ether was added to remaining residue. The traces of pyridine was
extracted with 5% HCl 2ꢂ 10 mL. Fluorocitric acid diethyl ester was
extracted out with 5% NaHCO3 solution 2ꢂ 20 mL. The water
solution was cooled (ice bath), ether (75 mL) was added, and then
it was acidified with 0.075 mol of 10% HCl. The organic phase was
separated, dried with MgSO4. After removal of ether 13.7 g (69%) of
pure fluorocitric acid diethyl ester 6 was obtained. All attempts to
separated and dried with MgSO4. After removal of ether 2.7 g (69%)
of pure dibenzyl 2-fluoro-3-oxosuccinate 8 was obtained.
4.6.2. Spectroscopic data for
D
,
L
-fluorocitric acid dibenzyl ester (8)
1H NMR (600 MHz, CDCl3)
d
7.42–7.20 (m, 10H), 5.15 (d,
13.5 Hz, 1H), 5.12 (d, 47.1 Hz, 1H), 5.12 (d, 11.3 Hz, 1H), 5.10 (d,
11.9 Hz, 1H), 5.04 (d, 11.9 Hz, 1H), 3.07 (d, 16.5 Hz, 1H), 3.02 (d,
16.5 Hz, 1H). 13C NMR (151 MHz, CDCl3)
d 174.04 (s), 170.57 (d,
6.2 Hz), 165.82 (d, 25.0 Hz), 134.48 (s), 134.39 (s), 128.89 (s),
128.87 (s), 128.84 (s), 128.75 (s), 128.73 (s), 128.68 (s), 90.88 (d,
198.3 Hz), 76.25 (d, 20.0 Hz), 68.97 (s), 68.00 (s). EI-MS: m/z 381;
HRMS calcd for C23H29N2O3: 381.2178; found (ESI, [M+H]+):
381.2183.
obtain the free D,L-fluorocitric acid 5 by acidic (HCl, H2SO4) or basic
(NaOH, KOH) hydrolysis varying the temperature and time regimes
of both the triethyl and diethyl esters of fluorocitric acids failed.
4.4.2. Spectroscopic data for
1H NMR (600 MHz, CDCl3)
4.33–4.21 (m, 4H), 3.09 (d, 4.8 Hz, 1H), 3.04 (s, 1H), 1.31 (t, 7.2 Hz,
3H), 1.29 (t, 7.1 Hz, 3H). 13C NMR (151 MHz, CDCl3)
174.31 (s),
D
,
L
-fluorocitric acid diethyl ester (6)
6.94 (bs, 2H), 5.09 (d, 47.3 Hz, 1H),
4.7. Synthesis of D,L-fluorocitric acid tribenzyl ester (9)
d
4.7.1. General procedure for synthesis of D,L-fluorocitric acid tribenzyl
d
ester (9)
170.74 (d, 6.4 Hz), 166.06 (d, 24.4 Hz), 90.96 (d, 198.2 Hz), 76.11 (d,
19.9 Hz), 63.41 (s), 62.40 (s), 39.48 (d, 6.2 Hz), 14.08 (s), 13.99 (s).
Dibenzyl 2-fluoro-3-oxosuccinate 7 (3.3 g, 0.01 mol) was added
over a period of 30 min to a solution of malonic acid monobenzyl
ester [5] 1.94 g, (0.01 mol) in 5 mL of pyridine. The reaction
mixture was stirred at room temperature for 2 h, then heated on a
steam bath for ꢀ30 min until evolution of CO2 ceased. Pyridine was
evaporated, and 25 mL of ether was added to remaining residue.
The traces of pyridine was extracted with 5% HCl (2ꢂ 5 mL), excess
of malonic acid monobenzyl ester washed out with 5% NaHCO3
solution (2ꢂ 5 mL), water. This ether solution was dried with
MgSO4. After removal of ether 3.1 g (64.6%) of pure tribenzyl
monofluorocitrate 9 was obtained.
4.5. Synthesis of dibenzyl 2-fluoro-3-oxosuccinate (7)
4.5.1. General procedure for synthesis of dibenzyl 2-fluoro-3-
oxosuccinate (7)
To a cooled (ice bath) stirred solution of 0.0625 mol of sodium
hydride, in 100 mL of dry toluene was added dibenzyl oxalate
13.5 g (0.05 mol) and after stirring for 15 min solution of 8.4 g
(0.05 mol) of benzyl fluoroacetate2 [4] in 15 mL of dry toluene was
added dropwise. The mixture was stirred an additional 2–3 h
allowed to come to room temperature and then heated under
reflux for 15 min and left for a night at room temperature. The
solution was cooled the next day (ice bath) and to it was added
5 mL of ethanol to the mixture, and then, 0.0625 mol of 10% HCl
solution. The toluene layer was separated, dried on MgSO4 and
after evaporation of the toluene and benzyl alcohol the residue was
mixed with hexanes and cooled on dry ice. The solid material was
separated at room temperature and excess of hexanes was
4.7.2. Spectroscopic data for
D
,
L
-fluorocitric acid tribenzyl ester (9)
1H NMR (600 MHz, CDCl3)
d
7.37–7.24 (m, 15H), 5.13 (d,
12.1 Hz, 1H), 5.11 (d, 47.0 Hz, 1H), 5.09 (d, 12.1 Hz, 1H), 5.04 (s,
2H), 5.03 (d, 14.4 Hz, 1H), 4.98 (d, 11.9 Hz, 1H), 3.07 (d, 16.3 Hz,
1H), 3.01 (d, 16.2 Hz, 1H). 13C NMR (151 MHz, CDCl3)
d 170.51 (d,
6.4 Hz), 169.14 (s), 165.73 (d, 25.0 Hz), 135.15 (s), 134.42 (s),
134.39 (s), 128.65 (s), 128.62 (s), 128.59 (s), 128.57 (s), 128.56 (s),
128.52 (s), 128.39 (s), 128.32 (s), 90.82 (d, 197.9 Hz), 76.31 (d,
19.9 Hz), 68.59 (s), 67.70 (s), 66.88 (s), 39.68 (d, 5.7 Hz). [M+H]+,
calculated: 480.1662; measured: 480.1673 (1.3 ppm error).
evaporated giving 15.05 g (91.2%) of mixture of D,L-dibenzyl 2-
fluoro-3-oxosuccinate 7. Attempts to distill the product led to
decomposition.
4.8. Synthesis of D,L-fluorocitric acid (5)
4.5.2. Spectroscopic data for dibenzyl 2-fluoro-3-oxosuccinate (7)
1H NMR (600 MHz, CDCl3)
d
7.46–7.06 (m, 10H), 5.23 (d,
4.8.1. General procedure for synthesis of D,L-fluorocitric acid (5) from
9
48.1 Hz, 1H), 5.19–4.93 (m, 4H). 13C NMR (151 MHz, CDCl3)
d
168.49 (s), 167.94 (d, 2.4 Hz), 166.14 (d, 24.0 Hz), 136.95 (s),
136.61 (s), 128.67 (s), 128.61 (s), 128.52 (s), 128.30 (s), 128.27 (s),
88.61 (d, 188.1 Hz), 68.75 (s), 68.60 (s).
3.0 g of tribenzyl monofluorocitrate 9 in 60 mL of ethanol was
hydrogenated at the presence of 0.3 g of 10% palladium on charcoal
catalyst at 50 psi pressure and at room temperature for 24 h. The
catalyst was filtered off and the filtrate evaporated to dryness in
4.6. Synthesis of
D,
L-fluorocitric acid dibenzyl ester (8)
vacuo. The residue consisted of 1.3 g (100% yield) of D,L-fluorocitric
acid 5 as a colorless oil which partly crystallizes on standing during
4.6.1. General procedure for synthesis of
D,
L-fluorocitric acid dibenzyl
a long time.
ester (8)
Dibenzyl 2-fluoro-3-oxosuccinate 7 (3.3 g, 0.01 mol) was added
over a period of 30 min to a solution of malonic acid 1.04 g,
(0.01 mol) in 5 mL of pyridine. The reaction mixture was stirred at
room temperature for 2 h, then heated on a steam bath for ꢀ30 min
until evolution of CO2 ceased. Pyridine was evaporated, and 50 mL
of ether was added to remaining residue. The traces of pyridine
were extracted with 5% HCl (2ꢂ 5 mL), and the solution washed
with water. To the organic part was added 0.02 mol of NaHCO3 as
5% solution. The ether layer was separated, and to the cooled (ice
bath) alkali water suspension was added a new portion of ether
(50 mL) and it was acidified with 10% H2SO4. The organic layer was
4.8.2. General procedure for synthesis of D,L-fluorocitric acid (5) from
8
2.6 g of dibenzyl monofluorocitrate 8 in 52 mL of ethanol was
hydrogenated at the presence of 0.26 g of 10% palladium on
charcoal catalyst at 50 psi pressure and at room temperature for
24 h. The catalyst was filtered off and the filtrate evaporated to
dryness in vacuo. The residue consisted of 1.3 g (100% yield) of D,L-
fluorocitric acid 5 as a colorless oil which partly crystallizes on
standing during a long time.
4.8.3. Spectroscopic data for
1H NMR (600 MHz, CD3OD)
16.1 Hz, 1H), 2.91 (dd, 16.0, 0.9 Hz, 1H). 13C NMR (151 MHz,
D,
L-fluorocitric acid (5)
d
5.20 (d, 47.4 Hz, 1H), 3.05 (d,
2
Caution: Benzyl fluoroacetate is a highly toxic compound.

924
R. Vardanyan et al. / Journal of Fluorine Chemistry 132 (2011) 920–924
[7] E.D. Milligan, L.R. Watkins, Nat. Rev. Neurosci. 10 (2009) 23.
CD3OD)
d 173.73 (d, 6.5 Hz), 172.87 (s), 169.68 (d, 25.8 Hz), 92.71
[8] M. Takeda, M.I. Takahashi, S. Matsumoto, Neurosci. Biobehav. Rev. 33 (2009) 784.
[9] L.R. Watkins, M.R. Hutchinson, A. Ledeboer, J. Wieseler-Frank, E.D. Milligan, S.F.
Maier, Brain Behav. Immun. 21 (2007) 131.
(dd, 193.2, 5.8 Hz), 77.40 (d, 20.2 Hz), 40.91 (d, 6.2 Hz). 19F NMR
(282 MHz, CD3OD)
d
ꢁ78.5 (s, TFA), ꢁ202.73 (d, 47.5 Hz).
[10] M.R. Suter, Y.R. Wen, I. Decosterd, R.R. Ji, Neuron Glia Biol. 3 (2007) 255.
[11] J. Scholz, C.J. Woolf, Nat. Neurosci. 10 (2007) 1361.
[12] J. Mika, Pharmacol. Rep. 60 (2008) 297.
Acknowledgments
[13] L.R. Watkins, S.F. Maier, Drug Discov. Today: Ther. Strat. 1 (2004) 83.
[14] M.R. Hutchinson, S.T. Bland, K.W. Johnson, K.C. Rice, S.F. Maier, L.R. Watkins, Sci.
World J. 7 (S2) (2007) 98.
[15] L.R. Watkins, M.R. Hutchinson, E.D. Milligan, S.F. Maier, Brain Res. Rev. 56 (2007)
148.
This work was supported by grants from the U.S. Public Health
Service, National Institute of Health, DA 13449 and DA 06284.
References
[16] L.R. Watkins, S.F. Maier, in: H. Merskey, J.D. Loeser, R. Dubner (Eds.), Paths of Pain,
IASP Press, Seattle, WA, 2005, p. 165.
[17] T. Zhang, J. Zhang, J. Shi, Y. Feng, Z.S. Sun, H. Li, Brain Res. Bull. 79 (2009) 219.
[18] J. Scholz, C.J. Woolf, Nat. Neurosci. 5 (S) (2002) 1062.
[19] D.E.A. Rivett, J. Chem. Soc. (1953) 3710.
[20] P.F.V. Ward, E.M. Gal, R.A. Peters, Biochim. Biophys. Acta 21 (1956) 392.
[21] O. Lerman, S. Rozen, J. Org. Chem. 48 (1983) 724.
[22] I. Shahak, E.D. Bergmann, J. Chem. Soc. (1960) 3225.
[23] A.B. Shtarev, E. Pinkhassik, M.D. Levin, I. Stibor, J. Michl, J. Am. Chem. Soc. 123
(2001) 3484.
[1] R. Morphy, Z. Rankovic, J. Med. Chem. 48 (2005) 6523.
[2] R. Morphy, C. Kay, Z. Rankovic, Drug Discov. Today 9 (2004) 641.
[3] R. Morphy, Z. Rankovic, Drug Discov. Today 12 (2007) 156.
[4] D. Vance, M. Shah, A. Joshi, R.S. Kane, Biotechnol. Bioeng. 101 (2008) 429.
[5] S. Ballet, M. Pietsch, A.D. Abell, Protein Peptide Lett. 15 (2008) 668.
[6] V.J. Hruby, F. Porreca, H.I. Yamamura, G. Tollin, R.S. Agnes, Y.S. Lee, M. Cai, C.S.
Alves, E. Varga, P. Davis, Z. Salamon, W. Roeske, T. Vanderah, J. Lai, Am. Assoc.
Pharm. Sci. J. 8 (2006) E450.