Facile synthesis of high specific activity 4-[1-14C]butyl-1,2-diphenylpyrazolidine-3,5-dione (phenylbutazone) using nucleophilic substitution

Article abstract of DOI:10.1002/jlcr.3597

Metabolism, environmental fate, and low concentration food residue studies would be facilitated by the use of radiolabeled test articles that can be readily quantified within complex matrices. However, radiochemical approaches for such studies require high specific activities to allow analytical detection of correspondingly low masses of test article. The synthesis of high specific activity (>50?μCi/μmol) [¹⁴C]-radiolabeled phenylbutazone presents a challenge using existing methodology, mainly due to the low solvent volumes required and vigorous refluxing needed to close the pyrazolidinedione ring. Herein, we report on the significant modification of an existing method that allows the synthesis of low masses of high specific activity (>50?μCi/μmol) [¹⁴C]-phenylbutazone under mild conditions with simple purification and high yield. The closure of the pyrazolidinedione ring of 1,2-diphenyl-3,5-pyrazolidinedione was accomplished as a first step with unlabeled 1,2-diphenylhydrazine and diethyl malonate in 32% yield under gram-scale conditions, which avoided the challenges of low solvent use and vigorous refluxing. Low mass of high specific activity n-[1-¹⁴C]-butyl bromide was then added via a nucleophilic substitution reaction as a final step. Yields ranged from 65% to 92% during multiple synthetic attempts with unlabeled butyl bromide and were greatly influenced by reaction stoichiometry and the selection of base.

Full text of DOI:10.1002/jlcr.3597

Facile synthesis of high specific activity 4-[1-14C]butyl-1,2-
diphenylpyrazolidine-3,5-dione (phenylbutazone) using nucleophilic
substitution
Anuradha Singh1, Heldur Hakk1, Sara J. Lupton1
1 USDA-ARS, Red River Valley Agricultural Research Center, Biosciences Research
Laboratory, Fargo, ND.
Correspondence:
Heldur Hakk, USDA-ARS Biosciences Research Laboratory, Fargo, ND 58102 USA
Email: heldur.hakk@ars.usda.gov
This article has been accepted for publication and undergone full peer review but has not
been through the copyediting, typesetting, pagination and proofreading process which may
lead to differences between this version and the Version of Record. Please cite this article as
doi: 10.1002/jlcr.3597
This article is protected by copyright. All rights reserved.

Abstract
Metabolism, environmental fate, and low concentration food residue studies would be
facilitated by the use of radiolabeled test articles which can be readily quantified within
complex matrices. However, radiochemical approaches for such studies require high specific
activities to allow analytical detection of correspondingly low masses of test article. The
synthesis of high specific activity (>50 μCi/μmol) [14C] radiolabeled phenylbutazone presents
a challenge using existing methodology, mainly due to the low solvent volumes required and
vigorous refluxing needed to close the pyrazolidinedione ring. Herein we report on the
significant modification of an existing method that allows the synthesis of low masses of high
specific activity (>50 μCi/μmol) [14C]-phenylbutazone under mild conditions with simple
purification and high yield. The closure of the pyrazolidinedione ring of 1,2 diphenyl-3,5-
pyrazolidinedione was accomplished as a first step with unlabeled 1,2-diphenylhydrazine and
diethyl malonate in 32% yield under gram-scale conditions, which avoided the challenges of
low solvent use and vigorous refluxing. Low mass of high specific activity n-[1-14C]-butyl
bromide was then added via a nucleophilic substitution reaction as a final step. Yields ranged
from 65-92% during multiple synthetic attempts with unlabeled butyl bromide, and were
greatly influenced by reaction stoichiometry and the selection of base.
Keywords:
Phenylbutazone, anti-inflammatory, [14C]-radiolabel, butyl bromide, pyrazolidinedione,
nucleophilic substitution, stoichiometry, NSAID
This article is protected by copyright. All rights reserved.

1. Introduction
Phenylbutazone (PB; 4-butyl-1,2-diphenylpyrazolidine-3,5-dione) is a nonsteroidal
anti-inflammatory (NSAID) drug. It is widely used in animals, especially horses, to provide
relief from acute and chronic inflammatory pain.1,2 Because of concerns of related to food
safety, PB is not allowed for use in food animals. Investigations on the partitioning of PB
within biological fluids, its metabolism in animals, and laboratory-scale environmental fate
and transport are supported with high specific activity, radioisotopically-labeled material.
Subsequent use of liquid scintillation counting, tissue combustion analysis, or radiochemical
detection during chromatography are commonly used to quantitatively and qualitatively
assess chemical fate and disposition, but such experiments are facilitated by high specific
activity test articles, especially when low masses of analyte are encountered (as in tissue
residue experiments). PB is typically produced by a condensation reaction between n-butyl-
diethylmalonate and 1,2-diphenylhydrazine in the presence of bases such as sodium ethoxide,
potassium t-butoxide, or sodium hydride in 40-65% yield at high temperatures.3-5 This
reaction is performed on a reasonably large (milligram or gram) scale. The economical
synthesis of high specific activity, radiolabeled PB requires microgram scale reactions using
limited solvent, while limiting the consumption of expensive radiolabeled reagent.
To the best of our knowledge there is no literature that describes a methodology for
synthesizing high specific activity [14C]-phenylbutazone (>50 μCi/μmol) at the microgram
scale. Ayrey and Yeomans5 demonstrated the synthesis of low specific activity PB (~0.3
μCi/μmol) using a reasonably high mass of reagents (e.g. high milligram scale). The
synthesis was accomplished in one-pot with a multi-step temperature gradient via the
condensation of [14C]-diethyl malonate (1:20 ratio of radiochemical:unlabeled compound),
butyl bromide, and 1,2-diphenylhydrazine in the presence of sodium ethoxide (Fig. 1). The
This article is protected by copyright. All rights reserved.

reaction required a stoichiometric amount of moisture-sensitive base, that suffered from
highly variable yields if any residual water from solvents, reagents, or glassware were
present; a highly detrimental condition when attempting micromolar scale synthesis of a high
specific activity product. We were not able to adapt the method of Ayrey and Yeomans5 to the
micromolar scale to accomplish the synthesis of high specific activity radiolabeled PB since
direct scale-down required low solvent volume (<5 μL), ultra-dry conditions, three stages of
vigorous heating, as well as a tightly controlled stoichiometric amount of sodium ethoxide
base. Therefore, there was a need to discover a robust, reproducible, and high-yielding
reaction to make radiolabeled PB which required low quantities of solvent and radiolabeled
reagents. Herein a simple, robust, reproducible, and high-yielding methodology to synthesize
radiolabeled PB is described.
2. Result and Discussion
The initial reaction attempt was conducted at the millimolar scale with unlabeled
reagents according to the published method of Ayrey and Yeomans.5 An equimolar (4.74
mmol) condensation of diethyl malonate, butyl bromide, and 1,2-diphenylhydrazine was
performed in the presence of sodium ethoxide and absolute ethanol with stepwise heating to
150 oC.5 The reaction had very acceptable yields (71%) at the millimolar scale. However,
for the high-specific activity radiolabeled synthesis of PB, the original goal was to perform
the reaction with ~5 µmol (800 µg; 250 μCi) of radiolabeled diethyl malonate. Thus, a series
of reactions that varied the amount of diethyl malonate, ranging from 4.74 mmol to 0.15
mmol, was performed to investigate suitability of the reaction at these lower scales (data not
shown). Phenylbutazone product was observed in reactions that used diethyl malonate at
amounts greater than 0.59 mmol, but product was not observed when diethyl malonate
amounts were below 0.3 mmol. Yields were 30-50% when diethyl malonate was greater than
This article is protected by copyright. All rights reserved.

0.59 mmol as assessed by gas chromatography-mass spectrometry (GC-MS). Further
reactions were conducted at 0.15 mmol diethyl malonate with different solvents (toluene,
chlorobenzene, and xylene) to investigate solvent influences on the condensation reaction;
however, no product was detected by GC-MS. Condensation reactions between pre-alkylated
butyl diethyl malonate (0.1 mmol) and 1,2-diphenylhydrazine in the presence of sodium
ethoxide or potassium tert-butoxide bases were also unsuccessful.6 Thus, another approach
was developed to make high specific activity [14C]-PB at a small scale with reasonably high
chemical yield. One element of the new approach involved introducing the radiolabel at the
final synthetic step rather than during an initial reaction step. As a consequence, radiolabeled
diethyl malonate was replaced with radiolabeled butyl bromide (Fig. 2). An additional
element was maintaining an optimized stoichiometry for all reagents in the final reaction step.
Finally, use of a base, i.e. potassium carbonate, that could be gravimetrically measured with
high accuracy replaced sodium ethoxide, mentioned in the published literature, but which
suffered from variable base strength.
Since results from the initial millimolar scale reactions indicated that cyclization was the
rate-limiting step, gram-scale synthesis of 1,2- diphenyl-3,5-pyrazolidinedione (3) was
performed first, and resulted in the facile production of desired product (32%). Thus,
sufficient amounts of a rate-limiting intermediate (3) were available to facilitate numerous
unlabeled reactions at the scale needed for radiolabeled butyl bromide introduction via
nucleophilic substitution (Fig. 2). 1,2 diphenyl-3,5-pyrazolidinedione (3) was synthesized via
condensation of 1,2-diphenylhydrazine (1) and diethyl malonate (2) in the presence of
sodium hydride base in chlorobenzene.7 Improvements to the published method5 included
introduction of sodium hydride base, a solid which resulted in high confidence in the base
stoichiometry, and the use of chlorobenzene as solvent, which was less hygroscopic than
ethanol, and which possessed a higher boiling point, facilitating the first step of ring closure.
This article is protected by copyright. All rights reserved.

Alkylation of 2.4 equivalents of 3 with 1 equivalent of butyl bromide (4; 5 μmol; 685 μg;
0.537 μL) in the presence of 2 equivalents of potassium carbonate base and a catalytic
amount of sodium iodide yielded 60% (5) by GC-MS (Reaction 1, Table 1), and 30%
dibutylated by-product. Incomplete and variable formation of sodium ethoxide from sodium
metal and absolute ethanol was detrimental, and contributed to lower yields of (5). Solid
potassium carbonate was a more suitable base than liquid sodium ethoxide used by Ayrey and
Yeomans,5 and the higher molecular weight and lower hygroscopic properties made it
preferable over sodium hydride in this low mass synthetic step. Subsequently, the
monoalkylation reaction was optimized by varying the equivalents of K2CO3 and (3) (Table
1). The scale for optimizing the reaction was chosen on the basis of availability of radiolabel
butyl bromide (5 μmol, ~250 μCi). Acetonitrile was a preferred solvent for the nucleophilic
substitution reaction because of its polar character and ability to be dried adequately.
However, the reaction was also run with hexane since the radiolabeled butyl bromide was
packaged in hexane. During nitrogen flushing at 0°C biphasic conditions occurred, but a
single phase existed when heating to 55°C during the reaction. Knovenagel condensation
with 1,2-diphenyl-3,5-pyrazolidinedione (3) and butanal followed by hydrogenation could
also have produced (5).8 However, this approach was not followed since it required two
additional steps with putative losses, obviously undesirable once the radiolabel had been
introduced.
As shown in Table 1, use of 2.4 equiv of 3 resulted in good conversion but with more
dibutylation, whereas the use of 6 equiv of 3 slowed the reaction and resulted in low yield at
24 h. The optimized reaction 3 conditions (Table 1) produced the highest yield of 5, and were
repeated seven times with unlabeled reagents with yields that ranged from 65-92% of PB (5)
by GC-MS. Pure 5 was produced for NMR analysis by purification with silica gel thin layer
chromatography (TLC). Optimized reaction 3 conditions were applied to radiolabeled PB
This article is protected by copyright. All rights reserved.

synthesis utilizing n-[1-14C]-butyl bromide, which was directly used from the ampule in
which it was shipped (250 µCi in 1 ml hexane). No attempt to evaporate the hexane (BP
68.5°C) was pursued due to the reasonably volatile nature of butyl bromide (BP 101.4°C).
After the extraction of the radiolabeled reaction with dichloromethane, the organic layer was
dried over sodium sulfate, and crude radiochemical (83.7 µCi ) was purified by prep TLC
developed in 1:1 hexane:ether. Radiolabeled 5 was obtained in 9% yield (Rf = 0.3; 23 μCi;
>98% pure; 57 μCi/μmol), which was confirmed by GC-MS. A significant amount of
dibutylation unexpectedly occurred as the main by-product (21.7 μCi; Rf = 0.6, 9% yield;
4,4-dibutyl-1,2-diphenylpyrazolidine-3,5-dione). An uncharacterized oxidation product 5,9,10
was also observed near the origin (Rf = 0.04; 23.3 µCi, 9%). The aqueous layer contained
most of the remaining radioactivity, i.e. 166 μCi. The radiolabeled synthesis results were
negatively affected by unexpectedly high radiochemical content (656 μCi; 13 μmol) of n-[1-
14C]-butyl bromide. The stoichiometric excess (2.6-fold) of butyl bromide (13 vs 5 μmol) in
the commercial radiolabeled product caused a shift to dialkylation rather than the targeted
monobutylation. According to the reactions in Table 1, a shift from the optimal ratio of 4:1 (3
to butyl bromide) to this new ratio of 1.5:1 would favor dibutylation. For example, lowering
the optimal ratio to 2.4:1, as in reaction 1 (Table 1), resulted in a significant increase in
dibutylated by-product (15% vs. 30% for reaction 3). The lack of full mass balance of
radioactivity was probably due to the semi-volatile nature of n-[1-14C]-butyl bromide (BP
~100°C), which could be lost during reaction heating (55°C) or rotary evaporation of organic
extracts (45°C under vacuum). Nevertheless, the concept of synthesizing a low mass
phenylbutazone with high [14C] specific activity needed to conduct future food fate studies in
our laboratory was proven.
3. Conclusion
This article is protected by copyright. All rights reserved.

A simple, mild, robust, relatively high-yielding method to prepare radiolabeled
phenylbutazone was developed. It is recommended that the exact stoichiometry of all
reagents developed in these studies is used for highest yield, including K2CO3 base, but
especially butyl bromide so that it will not be evaporated under the positive nitrogen pressure
maintained in this reaction, or result in the formation of the undesirable dibutylated product.
4. Experimental
Diethylmalonate, butyl bromide, 1,2-diphenylhydrazine were purchased from Sigma-
Aldrich (St. Louis, MO) and used without further purification. Acetonitrile was distilled over
CaH2 and stored over 4Å molecular sieves (Linde). Radiolabeled n-[1-14C]-butyl bromide
was purchased from American Radiolabeled Chemicals (57 μCi/μmol; 250 μCi; St. Louis,
MO). A radioactive assay of the small amount of n-[1-14C]-butyl bromide remaining in the
shipped ampoule after its use in the reaction yielded a considerably higher amount of actual
material received, i.e. 656 μCi, 13 μmol; 1.8 mg, which was 2.6-fold more than stated. NMR
spectra were obtained on a 400 MHz Bruker BZH400/89 spectrometer. For GC-MS, a
Hewlett Packard 5890 with auto sampler (75-300 oC, 5 min solvent delay, temperature
gradient 10 o/min, final hold 10 min) was used. Prep TLC purification of PB was performed
using silica gel GF plate (Analtech, Newark, DE; 250 µm, 5 x 20 cm), and was visualized
with ultraviolet light (280 nm; for non-radiolabeled detection) or a System 2000 Imaging
Scanner (Bioscan, Washington, DC; radiochemical detection).
Synthesis of 1, 2-diphenyl-3, 5-pyrazolidinedione (3):
1,2-Diphenyl-3,5-pyrazolidinedione (3) was synthesized according to an published method.3b
To a mixture of 1,2-diphenylhydrazine (1; 1.84 g, 10 mmol), diethyl malonate (2; 1.6 g, 10
mmol) and chlorobenzene (8 mL), sodium hydride (60% mineral oil; 480 mg, 12 mmol) was
added in three portions at 0 oC. Reaction was refluxed for 5 h then cooled to room
This article is protected by copyright. All rights reserved.

temperature. Water (20 mL) was added to the flask to quench the reaction. The reaction was
extracted with dichloromethane (4 x 20 mL). The aqueous layer was acidified with 2 M HCl
to pH 2 and then extracted with dichloromethane (3 x 20 mL). Combined organic layers were
dried over sodium sulfate and concentrated in a rotary evaporator to obtain 1.2 g of crude 1,2-
diphenyl-3,5-pyrazolidinedione (3) in 32 % yield (0.8 g, 3.2 mmol). Sample was
recrystallized from absolute ethanol. Product was confirmed by 1H-, 13C-NMR and GC-MS.
1H-NMR (CDCl3, 400 MHz): δ 3.57 (s, 2H), 7.18-7.37 (m, 10H); 13C-NMR (CDCl3, 100
MHz): δ 38.7, 124.9, 129.3, 131.3, 137.5, 168.7. GC-MS (m/z, relative intensity) 252 (M+,
50), 183 (58), 105 (21), 77 (100).
Synthesis of non-radiolabeled phenylbutazone (5):
To a 10 mL vial, 1,2-diphenyl-3,5-pyrazolidinedione (3, 50 mg, 200 μmol, 4 equiv),
potassium carbonate (10 mg, 75 μmol, 1.5 equiv), sodium iodide (single crystal), butyl
bromide (4, 7 mg, 5.5 μL, 50 μmol, 1 equiv), acetonitrile (5 mL) and hexane (2.5 mL) were
added. The reaction vial was placed into an ice bath while nitrogen was bubbled into the vial
for 15 minutes with an exit needle to exclude oxygen. Then the exit needle was removed and
the nitrogen inlet was replaced by N2 balloon and sealed to the vial with Parafilm®. The
reaction was removed from the ice bath and heated at 55 oC for 22 h, after which time the
reaction was cooled to room temperature. The solvent was removed via rotary evaporation.
Water (5 mL) was added to the residue and extracted with dichloromethane (4 mL x 3).
Combined organic layers were dried over sodium sulfate and the organic layer was
concentrated by a stream of nitrogen, leaving crude PB (5), which was purified with TLC
1
developed with 1:1 hexane:ether (12 mg, 38.9 μmol; 78% yield). H- and 13C-NMR11, and
GC-MS12 matched with literature values. 1H-NMR (CDCl3, 400 MHz): δ 7.4-7.1 (m, 10H),
3.40 (t, J=8 Hz, 1H), 2.12-2.10 (m, 2H), 1.53-1.43 (m, 2H), 1.42-1.31(m, 2H), 0.90 (t, J = 8
This article is protected by copyright. All rights reserved.

Hz, 3H). 13C-NMR (CDCl3, 100 MHz): δ 170.6, 135.9, 127.0, 122.7, 46.5, 28.1, 24.4, 22.3,
13.9. GC-MS (m/z, relative intensity) 308 (M+, 35), 252 (9), 183 (80), 105 (16), 77 (100).
Synthesis of [14C]radiolabelled phenylbutazone (5):
To an ice-cold 4 mL vial 1,2-diphenyl-3,5-pyrazolidinedione (3, 4.4 mg, 17.4 μmol, 4 equiv),
potassium carbonate (1.2 mg, 8.7 μmol, 2 equiv), sodium iodide (single crystal), n-[1-14C]-
butyl bromide in 0.25 mL hexane (assumed 596 mg, 4.35 μmol, 1 equiv, ~250 μCi),
acetonitrile (0.5 mL) were added. Reaction was flushed with nitrogen for 15 minutes over
icebath. The reaction was heated to 55 oC for 23 h, and allowed to cool to room temperature.
Water (5 mL) was added to quench the reaction and then the contents were extracted with
dichloromethane (3 mL x 4). Combined organic layers were dried over sodium sulfate, and
organic layer was concentrated by a stream of nitrogen. Radiolabeled phenylbutazone was
purified by TLC with 1:1 hexane:ether development. Characterization of radiolabeled
phenylbutazone was confirmed by GC-MS. GC-MS (m/z, relative intensity) 310 (M+, 27),
252 (8), 183 (84), 105 (16), 77 (100).
Acknowledgments
The use of trade, firm, or corporation names in this publication is for the information and
convenience of the reader. Such use does not constitute an official endorsement or
approval by the United States Department of Agriculture and the Agricultural Research
Service of any product or service to the exclusion of others that may be suitable. USDA is
an equal opportunity provider, employer, and lender. The authors declare no competing
financial interest.
This article is protected by copyright. All rights reserved.

References
1. Lees P, Toutain PL. Pharmacokinetics, pharmacodynamics, metabolism, toxicology
and residues of phenylbutazone in humans and horses. Vet J 2013;196: 294-303.
2. Tobin T, Chay S, Kamerling S, Woods WE, Weckman TJ, Blake JW, Lees P.
Phenylbutazone in the horse: A review. J Vet Pharmacol Therap 1986;9:1-25.
3. Gunther H, Ingo R, Ulf F. New syntheses of pyrazolidine and hexahydropyridazine
derivatives. Chem Berich 1957;90:537-42
4. Khaletskii AM, Pesin VG, Syan DZ. Chemistry of pyrazolidine. I. Synthesis and study
of some mono- and disubstituted 1,2-diphenyl-3,5-dioxopyrazolidines. Russ J Gen
Chem 1958;28:2355-2359.
5. Ayrey G, Yeomans MA. Synthesis and analysis of [4-14 C]-phenylbutazone. J
Radioanal Chem 1974;20:463-471.
6. Kende AS, Koch K, Smith CA. Cyclization of phenolic enolates J Amer Chem Soc
1988;110:2210-2218.
7. Vennerstorm J L, Holmes TJ Jr. Preparation and evaluation of electrophilic derivatives
of phenylbutazone as inhibitors of prostaglandin-H-synthase. J Med Chem
1987;30:563-567.
8. Ursula W. On the behaviour of some derivatives of 5-pyrazolone and 3,5-
pyrazolidinedione towards sodium borohydride. Justus Liebigs Ann Chem
1975;12:2293-2304.
9. Stig V, Linholter S, Sørensen P. Pyrazole studies: XI. Oxidation by air of 1,4 –
disubstituted pyrazolidine 3,5-diones. Acta Chem Scand 1958;12: 1359-1363.
10. Awang DVC, Vincent A, Matsui F. Pattern of phenylbutazone degradation. J Pharm
Sci 1973;62:1673-1676.
This article is protected by copyright. All rights reserved.

11. Sigma-Aldrich® website:
http://www.sigmaaldrich.com/catalog/product/sigma/p8386?lang=en&region=US
(accessed 4/17/17).
12. Neto LMR, Andraus MH, Salvadori MC. Determination of phenylbutazone and
oxyphenbutazone in plasma and urine samples of horses by high-performance liquid
chromatography and gas chromatography-mass spectrometry. J Chrom B
1996;678:211-218.
This article is protected by copyright. All rights reserved.

Table 1. Optimization of reaction conditions for the final step in the synthesis of
[12C]phenylbutazone (5) utilizing 1, 2-diphenyl-3, 5-pyrazolidinedione (3), potassium
carbonate base, and butyl bromide. Reactions were conducted with 1 equivalent of [12C]butyl
bromide (5 μmol) in 0.5 mL acetonitrile and 0.25 mL hexane.
Reaction #
Equiv. K2CO3
Equiv. 3
Time (h)
% (5)
%
Dibutylation
1
2
3
4
5
2
2.4
6
22
24
18
6
60
3
30
1
2
2
4
85 a
15 a
0
2
4
20
66
1.5
4
18
4
a Value is the mean of seven reactions.
This article is protected by copyright. All rights reserved.

Figure 1. Synthesis of [14C]-radiolabeled phenylbutazone by the one step method of Ayrey
and Yeomans (1974). The method involved initial condensation of all three reagents at 70-
90°C, solvent evaporation and pyrazolidinedione ring closure at 120-150°C to yield
phenylbutazone.
This article is protected by copyright. All rights reserved.

Figure 2: Synthesis approach for [14C]-phenylbutazone (5); * indicates position of [14C]
labeled carbon. The first step involves formation of the unlabeled pyrazolidinedione ring (3)
by condensation of 1,2-diphenylhydrazine (1) and diethyl malonate (2). The introduction of
the radiolabel in the form of [14C]-butyl bromide (4) is performed as the final step in a
nucleophilic substitution reaction.
This article is protected by copyright. All rights reserved.