Easy preparative scale syntheses of labelled xanthines: Caffeine, theophylline and theobromine

Article abstract of DOI:10.1002/jlcr.1154

Several easy preparative scale (0.5-1.5 g) syntheses of deuterium labelled caffeine, theophylline and theobromine are described. Some new selective syntheses of theophylline and theobromine have been developed. Labelled xanthines are of great interest in qualitative or quantitative isotope dilution-mass spectrometry, coupled with gas or liquid chromatography, currently performed in anti-doping and forensic laboratories. Copyright

Full text of DOI:10.1002/jlcr.1154

Journal of Labelled Compounds and Radiopharmaceuticals
J Label Compd Radiopharm 2007; 50: 33–41.
Published online in Wiley InterScience
JLCR
(www.interscience.wiley.com). DOI: 10.1002/jlcr.1154
Research Article
Easy preparative scale syntheses of labelled xanthines:
Caffeine, theophylline and theobromine
´ ´
FREDERIC BALSSA* and YVES BONNAIRE
`
Laboratoire des Courses Hippiques, 15 rue de Paradis, 91370 Verrieres le Buisson, France
Received 20 September 2006; Revised 18 October 2006; Accepted 19 October 2006
Abstract: Several easy preparative scale (0.5–1.5 g) syntheses of deuterium labelled caffeine, theophylline and
theobromine are described. Some new selective syntheses of theophylline and theobromine have been developed.
Labelled xanthines are of great interest in qualitative or quantitative isotope dilution-mass spectrometry, coupled
with gas or liquid chromatography, currently performed in anti-doping and forensic laboratories. Copyright # 2007
John Wiley & Sons, Ltd.
Keywords: caffeine; theophylline; theobromine; labeled; deuterium
Introduction
unlabelled caffeine is at the molecular ion mass.1
Therefore, caffeine should be labelled at least on N-3
or N-7. Likewise, mass spectra of N-1 and N-3 labelled
theophylline2 indicated that the only difference be-
tween N-1 labelled and unlabelled theophylline was the
molecular ion mass, labelling of theophylline should be
performed at least on N-3. As theobromine exhibits a
similar fragmentation pattern, this compound should
also be labelled on N-3 or N-7.
The presence of natural xanthines (caffeine, theophyl-
line and theobromine, Figure 1) in biological samples is
regulated by International Racing Rules and/or Inter-
national Sports Authorities. In the horse, caffeine and
theophylline are prohibited and for theobromine, a
concentration threshold at 2.0 mg/ml in urine has been
established.
The use of labelled xanthines as internal standards is
recommended for analytical purposes, both in qualita-
tive screening and quantitative analysis in body fluids.
As no commercial source was available, it was decided
to prepare them. These labelled xanthines should be
chemically and isotopically pure (preferably a single
isotopomer), highly enriched, with none or trace
amounts of unlabelled compound.
Results and discussion
Caffeine
Two different syntheses of labelled caffeine have been
published: a total synthesis from [15N]urea3 and a
direct methylation of xanthine (or its methyl deriva-
tives) with 13CH3I4,5 or with CD3I.6–9 This direct
methylation is obviously the most direct way to
caffeine, but a review of the literature highlights several
drawbacks: expensive starting materials4,6–8 use of a
large excess of labelled iodomethane,4,9 low to moder-
ate yields,6,7,9 reaction leading to a mixture of several
methylated xanthines difficult to separate,5–7 and
purification methods incompatible with large-scale
synthesis.5,6
As the internal standard may be exposed to strong
acid or base during the analytical procedure, labelling
should be made on non-exchangeable positions. In the
case of methylated xanthines, substitution of one or
more N-CH3 by N-CD3 is the most appropriate.
The mass spectra of labelled and unlabelled species
should exhibit the least possible common fragment
ions in order to reduce risk of interference during
analyses of biological matrices. The only difference
between the mass spectra of N-1 labelled and
Because only one equivalent of CH3I is theoretically
required to generate caffeine via N-7 methylation,
theophylline 1 was chosen as the starting material.
Among the specific N-7 alkylation methods,4,6,8,10–22
the RX/K2CO3/DMF system is the most popular and
*Correspondence to: Fre´de´ric Balssa, Laboratoire des Courses Hippi-
ques, 15 rue de Paradis, 91370 Verrie`res le Buisson, France.
E-mail: lch-fbalssa@wanadoo.fr
Copyright # 2007 John Wiley & Sons, Ltd.

34 F. BALSSA AND Y. BONNAIRE
Figure 1 Natural xanthines.
Figure 2 Synthesis of caffeine-d3.
seems to be a general one. However, the toxicity of DMF
and its high boiling point led us to develop an
alternative alkylating process. Caffeine-d3 2 was pre-
pared by alkylation of theophylline with CD3I (1.5
equivalent) in methanol in the presence of sodium
methylate (Figure 2).
published.20,24–27 Likewise, numerous syntheses of
1,3-disubstituted xanthine have been described.10–13,
28–30
The first procedure we tested is an improvement
of the historic Traube xanthine synthesis starting
from 6-aminouracil24 (Figure 3). Indeed, this com-
pound is inexpensive and, if needed, can easily be
prepared.24,31,32 The 5-nitroso derivative 5 was pre-
pared from 6-aminouracil 3 according to a classical
procedure14,24,30,33 and reduced to the corresponding
air-sensitive 1,3-dimethyl-5,6-diamino-uracil-d6 6 via
catalytic hydrogenation10,14,30 instead of the original
Zn/H3O+ reduction. Imidazole ring closure was per-
formed in one step33,34 instead of the original two step
procedure (formylation followed by alkaline ring
closure).35 This one step procedure using triethylortho-
formate (or formamide, dimethylethylacetate, diethox-
ymethylacetate) is now widely used for both xanthines
and purines.21,36,37 Unfortunately, the combined yields
of the two last steps were low (20%), leading to an
extensive loss of labelled material.
The disappearance of the 3H signal at 3.93 ppm
indicates complete labelling of the N-7 methyl, in good
agreement with published data.6 The mass spectrum of
caffeine-d3 exhibits a 3 amu shift for all characteristic
fragments, in good agreement with the fragmentation
pattern of labelled caffeine.1
Theophylline
Three synthetic methods for labelled theophylline have
been published. The first method proceeds via hydro-
gen/tritium exchange in presence of palladium leading
to low isotope enrichment (12%).23 The second method
proceeds via a N-7 pivaloyloxymethyl protection of 1-or
3-methylxanthine followed by alkylation of the unsub-
stituted nitrogen atom with CD3I and subsequent
removal of the protecting group.2 The third method
consists of the direct methylation of xanthine using
13CH3I/CH3I, leading to a mixture of mono-, di- and
trimethylxanthines, which were separated.5
In order to overcome low yield, we have developed a
second procedure starting from guanosine (Figure 4).
Guanosine
8 was easily converted into 7-benzyl-
xanthine 10.28 The classical CD3I/K2CO3/DMF proce-
dure was found to be the most efficient for the
conversion of 10 into 1,3-bis(trideuteromethyl)-7-ben-
zylxanthine 11. The alkylation of 10, using (CD3)2SO4
in a K2CO3/THF (or acetone) mixture, gave 11 with a
lower yield. All attempts to alkylate 10, using CD3I/
MeOH/MeONa (or K2CO3), led to a mixture of mono-
and dimethylated compounds (complete N-3 alkylation,
partial N-1 alkylation). Alkylation using (CD3)2SO4
in tetra-n-butylammonium fluoride/THF22 rapidly
Unfortunately, these methods are not suitable for
large-scale synthesis for the following reasons: low
yield,2 reactions leading to a mixture of substituted
xanthines and purification process incompatible with
large-scale preparation (preparative TLC2 and semi-
preparative liquid chromatography).5
Several selective syntheses of theophylline starting
from 6-amino-uracil derivatives have also been
Copyright # 2007 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2007; 50: 33–41
DOI: 10.1002.jlcr

EASY PREPARATIVE SCALE SYNTHESES OF LABELLED XANTHINES 35
Figure 3 Synthesis of theophyline-d6, method A.
Figure 4 Synthesis of theophyline-d6, method B.
gave complete dialkylation, however the ammonium
salt was difficult to separate from the alkylated
xanthine.
hazardous hydrogen gas and avoids chromatographic
purification.
The disappearance of 3H signals at 3.18 and
2.94 ppm indicates complete labelling of N-1 and N-3
methyl groups. The mass spectrum of the theophylline-
d6 exhibits a 6 amu shift of the molecular ion and a
3 amu shift of the main fragments in good agreement
with published data.1
Among the published N-3 or N-7 benzyl cleavage
procedures (AlCl3,18 Na/NH3,38 H2SO4/TFA/anisole,13,14
H2/Pd),18,28,29,38 catalytic low-pressure hydrogenation
was tested first and gave theophylline-d6 7 after
chromatographic purification. Debenzylation of 1,3-
dimethyl-7-benzylxanthine-d6 11 by catalytic transfer
hydrogenation39 gave, after recrystallization, pure
theophylline-d6 7. This first application of catalytic
hydrogen transfer to xanthine debenzylation was found
to be superior to the catalytic low-pressure hydrogena-
tion. This procedure gives a higher yield, avoids use of
Theobromine
The sole reported synthesis of labelled theobromine
consists in the direct alkylation of xanthine with CD3I
leading to a mixture of mono-, di-and trimethylated
Copyright # 2007 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2007; 50: 33–41
DOI: 10.1002.jlcr

36 F. BALSSA AND Y. BONNAIRE
xanthines, which are separated using semi-preparative
liquid chromatography.5 As this purification is not
convenient for large-scale preparation, we decided to
develop a new method suitable for preparative amounts
of theobromine-d3. The major drawback of the classical
synthetic method for theobromine (methylation of
xanthine or 3-methylxanthine)20,21,25 is that it affords
theobromine contaminated with small amounts of 3-
methylxanthine and caffeine (separation of 3-methyl-
xanthine and theobromine is difficult).
variations, leading to a quite low yield of theobromine-
d3 (515%).
The synthesis of theobromine-d6 was then attempted
via alkylation of the tris(trimethylsilyl)xanthine with
iodomethane at the N-3 and N-7 positions, while the
N-1 position remains unchanged.41 Xanthine was
readily silylated, by heating under reflux in hexam-
ethyldisilazane, to give tris(trimethylsilyl)xanthine as a
white solid. After dissolution in toluene and addition of
CD3I, the mixture was heated at 758C. GC/MS
monitoring established that 7-methylation was com-
pleted in less than 1 h. The second methylation at N-3
was completed after several days of heating and
required the use of up to 50 equivalent of CD3I. This
result led us to test the same procedure starting from 3-
methylxanthine instead of xanthine.
Several selective syntheses of theobromine have been
published: starting from adenine,38 from caffeine via N-
1 aminodemethylation40 and from xanthine via methy-
lation of its triethylsilyl derivative.41 Likewise, selective
syntheses of 3,7-xanthine derivatives have been re-
ported: from xanthine via its trimethylsilyl derivative42
and from 1-substituted-6-aminouracil.34
3-Methylxanthine 15 was readily prepared in three
steps via the stable hydrochloride salt of 1-methyl-5-6-
diaminouracil 14,43 silylated by heating under reflux in
hexamethyldisilazane and treated with CD3I (13
equivalents) to give theobromine-d3 17 (Figure 5).
Direct methylation of 3-methylxanthine was tested
using a CD3I/K2CO3/DMF mixture (with 1.5 equivalent
N-1 aminodemethylation of caffeine was the first
procedure tested because the preparation of caffeine-
d3 (N-7 labelled) is easy. The conversion of caffeine-d3
into 1-amino-3,7-dimethylxanthine-d3 and then into
theobromine-d3 was very disappointing in our hands;
the yield of aminodemethylation was subject to large
Figure 5 Synthesis of theobromine-d3, methods A and B.
Copyright # 2007 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2007; 50: 33–41
DOI: 10.1002.jlcr

EASY PREPARATIVE SCALE SYNTHESES OF LABELLED XANTHINES 37
Figure 6 Synthesis of theobromine-d3, method C.
of CD3I, in order to avoid any incomplete methylation of
3-methylxanthine), leading to a mixture of theobro-
mine-d3 17 and caffeine-d6 16 which was removed by
washing with chloroform.
Experimental
All reagents were obtained from Sigma-Aldrich (St.
Quentin Fallaviers, France) except guanosine hydrate
(Lancaster, Strasbourg, France) and were used without
further purification. NMR spectra were recorded on a
Bruker ARX250 instrument. GC/MS analysis were
performed using Agilent GC5890/MSD5972, equipped
with a 25 m DB-5MS column (J & W Scientific).
Since the four previous preparations of labelled
theobromine were only partially successful we decided
to develop a new synthetic method (Figure 6). The key
intermediate 22 (N-1 protected theobromine precursor)
was prepared by action of chloromethyl pivalate on
3-methyl-7-benzylxanthine 21. The pivaloyloxymethyl
protecting group was used because all attempts to
cleave 1-benzyltheobromine via catalytic hydrogena-
tion failed. After debenzylation and trideuteromethyla-
tion, 24 was deprotected by alkaline hydrolysis, leading
to theobromine-d3 17, which was easily purified by
sublimation under vacuum.
Caffeine-d3 2
CD3I (1 ml, 16 mmol) was added in one portion to a
solution of theophylline (1, 1.91 g, 10.7 mmol) and
sodium methylate (730 mg, 12.8 mmol) in methanol
(100 ml) at room temperature. After 4 days stirring in a
sealed flask in the dark, water (200 ml) was added and
the mixture was extracted with chloroform (2 Â 100 ml).
The organic fractions were combined, dried with brine
(100 ml) and anhydrous sodium sulfate. Evaporation of
chloroform gave 2 as a white solid. Caffeine-d3 was
The disappearance of the 3H signal at 3.54 ppm
indicates the complete labelling of the N-7 methyl. The
mass spectrum of the theobromine-d3 exhibits a 3 amu
shift of all the characteristic fragments in good
agreement with the fragmentation pattern of methy-
lated xanthine.1
Copyright # 2007 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2007; 50: 33–41
DOI: 10.1002.jlcr

38 F. BALSSA AND Y. BONNAIRE
dissolved in chloroform (30 ml), filtered and crystallized
by slow addition of diethyl ether. The caffeine-d3 2 was
collected by filtration, washed with diethyl ether and
dried under vacuum (1.31 g). The mother liquor was
evaporated, the remaining solid was dissolved in
chloroform (10 ml) and crystallized in the same way to
give 0.19 g of 2 (white crystals, combined: 1.50 g, 71%).
1H NMR (CDCl3) ppm: 7.60 (s, 1H), 3.52 (s, 3H), 3.34
(s, 3H). MS (EI), m/z (%): 197 (100), 168 (6), 139 (5),
112 (60), 85 (26), 70 (29), 67 (24), 58 (42), 45 (13).
Caffeine-d3 prepared from 250 mg of theophylline
(1.38 mmol), 97 mg of sodium methylate (1.66 mmol),
and CD3I (447 ml, 6.9 mmol) in methanol (15 ml) was
obtained after 24 h stirring at room temperature (white
crystals, 228 mg, 83.5%).
Theophylline-d6 7; method B
7-Benzylguanine Hydrochloride 9. Guanosine hydrate
8 (10 g, 33.2 mmol) was dissolved in dimethylsulfoxide
(50 ml) at 708C. After cooling at room temperature,
benzylbromide (10.0 ml, 84.1 mmol) was added in one
portion and the solution was stirred at room tempera-
ture in a sealed flask for 4 h. Concentrated hydrochloric
acid (37%, 25 ml) was added to the stirred solution in
one portion without cooling. After stirring for an
additional 1 h, the reaction mixture was poured into
well-stirred cold methanol (À108C, 300 ml). After an-
other 1 h stirring, the white crystalline precipitate was
collected by filtration, washed with cold methanol
(2 Â 100 ml) and dried under vacuum (white crystalline
solid, 5.01 g, 55.5%).
Theophylline-d6 7, method A
7-Benzylxanthine 10. 7-Benzylguanine hydrochloride 9
(5 g, 18 mmol) was dissolved in a mixture of acetic acid
(120 ml) and water (20 ml) at 1008C. After cooling to
508C, sodium nitrite (4.97 g, 72 mmol) in water (20 ml)
was added dropwise and stirring was continued for 4 h
at room temperature. The precipitate, formed upon
cooling, was collected by filtration, dried, washed with
diethyl ether (2 Â 100 ml) and dried under vacuum
(white solid, 3.0 g, 68.8%).
6-Amino-1,3-dimethyluracil-d6 4. (CD3)2SO4 (10 g,
75.7 mmol) was added in one portion to a suspension
of 6-aminouracil 3 (4.27 g, 33.6 mmol) in a mixture of
32% sodium hydroxyde solution (6.3 ml) and water
(26 ml), cooled in an ice bath. After 4.5 h stirring, the
solid was collected by filtration, washed with water
(30 ml) and dried (white solid, 2.53 g, 46.7%).
1,3-Dimethyl-7-benzylxanthine-d6 11. 7-Benzylxanthine
10 (400 mg, 1.65 mmol), potassium carbonate (920 mg,
6.7 mmol) and CD3I (400 ml, 6.43 mmol) in dry DMF
(5 ml) were stirred at 708C for 3 h in a sealed flask in the
dark. After cooling at room temperature, water (25 ml)
was added with stirring and the white precipitate was
collected by filtration, washed with water (2 Â 25 ml)
and dried under vacuum at 808C (white solid, 349 mg,
76.5%).
5-Nitroso-6-amino-1,3-dimethyluracil-d6 5. Sodium ni-
trite (430 mg, 6.2 mmol) in water (2 ml) was slowly
added to a stirred solution of 6-amino-1,3-dimethylur-
acil-d6 4 (500 mg, 3.1 mmol) in acetic acid (5 ml) and
water (5 ml), at 808C over a period of 1h. After stirring
for additional 30 min, the reaction mixture was cooled
in ice and stored overnight at 48C. The purple
precipitate of 5 was collected by filtration, washed with
cold water (10 ml) and dried over phosphorus pentoxide
(purple solid, 260 mg, 44%).
Theophylline-d6 7. Solid ammonium formate (1.5 g,
23.8 mmol) was added in one portion to a mixture of
1,3-dimethyl-7-benzylxanthine-d6 11 (660 mg, 2.39
mmol) and palladium on carbon (10%, 750 mg) in dry
methanol (50 ml) heated under reflux. After 3 h, the
palladium was removed by filtration through a celite
pad which was thoroughly washed with methanol
(100 ml). The solvent was removed by evaporation and
a tan solid was obtained. This solid was dissolved in hot
chloroform (10 ml) and filtered. After evaporation of
chloroform, the white solid was dissolved in methanol
(50 ml) and slowly cooled to À208C to yield (after 4 days)
white crystals (348 mg). The solid was filtered, washed
with cold methanol (À208C, 20 ml) and dried under
vacuum. The mother liquor was evapored, the collected
solid was dissolved in boiling methanol (8 ml) and
Theophylline-d6 7. A stream of hydrogen was passed
through a stirred mixture of 255 mg (1.34 mmol) of 5-
nitroso-6-amino-1,3-dimethyluracil-d6 5 and platinum
(IV) oxide (20 mg) in methanol (15 ml) using a fritted
pipe. After 3.5 h, the catalyst was removed by filtration
through a celite pad which was rinsed with dichlor-
omethane (100 ml). The filtrate was evaporated to give
5,6-diamino-1,3-dimethyluracil-d6 6 as a dark oil,
which was dissolved in triethylorthoformate (15 ml)
and heated under reflux for 4 h. Triethylorthoformate
was removed by distillation and the oily solid was
purified by chromatography on a silica gel column
(20 Â 3 cm, dichloromethane/methanol 2–10%) to give
theophylline-d6 7 (white solid, 51 mg, 20%).
Copyright # 2007 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2007; 50: 33–41
DOI: 10.1002.jlcr

EASY PREPARATIVE SCALE SYNTHESES OF LABELLED XANTHINES 39
crystallized similarly to give 79 mg of 7 (white crystals,
combined: 427 mg, 95.9%).
ether (2 Â 50 ml) and dried under vacuum at 908C
(white solid, 590 mg, 31%).
1H NMR (D2O / DCl) ppm: 8.44 (s, 1H). MS (EI), m/z
(%): 186 (100), 156 (7), 126 (6), 98 (68), 71 (45), 53 (32),
44 (20).
Theobromine-d3 17, method B
Theobromine-d3 17. 3-Methylxanthine 15 (440 mg,
2.66 mmol) was heated under reflux in hexamethyldi-
silazane (15 ml) under dry nitrogen until 3-methyl-
Theobromine-d3 17, method A
6-Amino-5-nitroso-1-methyluracil 13. A solution of
sodium nitrite (7.12g, 103.2 mmol) in water (40 ml) was
added dropwise to a stirred suspension of 6-amino-
1-methyluracil 12 (10.01 g, 70.9 mmol) in water (50 ml)
and acetic acid (20 ml) at room temperature. The
suspension was then heated at 508C for 1 h, stirred at
room temperature for 3 h and the purple solid was
collected by filtration. This solid was washed with ice-
water (20 ml), cold ethanol (100 ml) and dried under
vacuum at 908C (purple solid, 11.68 g, 99.8%).
xanthine
was
completely
dissolved
(24–48 h).
Hexamethyldisilazane was removed by evaporation
and the white crystals were suspended in dry toluene
(20 ml). CD3I (2.15 ml, 34.5 mmol) was added and the
reaction mixture was stirred at 758C in a sealed flask in
the dark for 3 days. After cooling at room temperature,
acetone (50 ml) was added with vigorous stirring,
followed by methanol (5 ml, dropwise). The precipitate
was collected by filtration, washed with acetone (50 ml),
dried under vacuum and crystallized from a minimum
volume of boiling water (white solid, 330 mg, 68%).
5,6-Diamino-1-methyluracil Hydrochloride 14. Solid so-
dium hydrosulfite (technical grade 85%, 20.01 g,
100 mmol) was slowly added to a suspension of 6-
amino-5-nitroso-1-methyluracil 13 (11.7 g, 68.8 mmol)
in water (150 ml). After complete addition, the mixture
was heated at 708C for 1 h, then cooled to room
temperature and filtered. The collected yellow solid
was washed with water (2 Â 100 ml), dried and added in
small portions to well stirred hot concentrated hydro-
chloric acid (37%, 50 ml, 908C). After stirring for an
additional 15 min, the suspension was cooled to room
temperature and filtered. The solid was washed with
acetone (100 ml) and dried under vacuum (white
crystalline solid, 6.15 g, 46.5%).
Theobromine-d3 17, method C
6-Amino-5-bromo-1-methyluracil 19. Sodium acetate
trihydrate (1.50 g, 11.0 mmol) was added to a suspen-
sion of 6-amino-1-methyluracil 12 (1.25 g, 8.8 mmol) in
acetic acid (40 ml) at 1258C. After cooling at 808C,
bromine (0.50 ml, 9.7 mmol) in acetic acid (5 ml) was
added dropwise to the suspension and heating was
stopped. The suspension was stirred for 48 h at room
temperature, the precipitate of 19 was collected by
filtration, washed with acetic acid (2 Â 25 ml) and dried
in vacuum at 908C (white solid, 1.94 g, 99.5%).
6-Amino-5-benzylamino-1-methyluracil 20. 6-Amino-
5-bromo-1-methyluracil 19 (11.0 g, 50 mmol) was sus-
pended in benzylamine (100 ml) at 1008C for 2 h. After
cooling at room temperature (precipitation occured),
ice-water (200 ml) was added with vigorous stirring and
the precipitate was collected by filtration. The solid 20
was washed with cold water (3 Â 100 ml) and dried in
vacuum at 908C (white solid, 7.63 g, 62%).
3-Methylxanthine 15. 5,6-Diamino-1-methyluracil hy-
drochloride 14 (6.15 g, 31.9 mmol) was heated at 708C
in a mixture of dry DMF (40 ml), triethylorthoformate
(13 ml, 71.8 mmol) and p-toluenesulfonic acid mono-
hydrate (50 mg) for 3.5 h (under nitrogen). After cooling
in an ice bath, water (100 ml) was slowly added with
vigorous stirring. The precipitate was collected by
filtration, washed with cold water (2 Â 100 ml), acetone
(100 ml) and dried under vacuum at 808C (white solid,
2.81 g, 53%).
3-Methyl-7-benzylxanthine 21. 6-Amino-5-benzylami-
no-1-methyluracil 20 (7.5 g, 30 mmol), triethylortho-
formate (14 ml, 77.3 mmol) and p-toluenesulfonic acid
monohydrate (150 mg) were heated at 708C for 48 h in
dry DMF (60 ml) under nitrogen. After cooling (ice
bath), ice-water (200 ml) was added and the suspension
was stirred for 15 min. The precipitate of 21 was
collected by filtration, washed with ice-water
(3 Â 100 ml), acetone (3 Â 100 ml), and dried in vacuum
at 508C (white solid, 5.60 g, 72 %).
Theobromine-d3 17. 3-Methylxanthine 15 (1.72 g,
10.3 mmol), potassium carbonate (3.0 g, 21.7 mmol)
and CD3I (1 ml, 16 mmol) in dry DMF (20 ml) were
heated at 808C for 2 h in a sealed flask in the dark. After
12 h at 48C, water (50 ml) was added and the solution
was neutralized with acetic acid. The precipitate was
collected by filtration, washed with water (2 Â 50 ml),
acetone (2 Â 50 ml), chloroform (2 Â 50 ml), diethyl
Copyright # 2007 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2007; 50: 33–41
DOI: 10.1002.jlcr

40 F. BALSSA AND Y. BONNAIRE
1-Pivaloyloxymethyl-7-benzyl-3-methylxanthine 22. A
mixture of 3-methyl-7-benzylxanthine 21 (5.5 g,
21.4 mmol), potassium carbonate (3.60 g, 26 mmol)
and chloromethyl pivalate (3.4 ml, 23.6 mmol) in dry
DMF (60 ml) was heated at 808C for 2 h under nitrogen.
After evaporation of the solvent, the yellow gum was
dissolved in chloroform (100 ml) and filtered through a
silica gel pad, yielding a yellow oil. This oil was
dissolved in diethyl ether (100 ml), hexane (50 ml) was
added with vigorous stirring (precipitation occurred)
and diethyl ether was slowly evaporated at room
temperature. The precipitate of 22 was collected by
filtration and dried (white solid, 6.97 g, 87.5%).
at 1208C (white solid, 590 mg, 68.5%). The purification
of theobromine-d3 was achieved by sublimation under
vacuum.
1H NMR (D2O / NaOD) ppm: 7.35 (s, 1H), 3.16 (s,
3H). MS (EI), m/z (%): 183 (100), 140 (6), 139 (6), 112
(29), 85 (28), 70 (39), 67 (25), 58 (45), 45 (11).
Conclusion
Simple and efficient preparations of labelled caffeine,
theophylline and theobromine have been perform-
ed, including new syntheses of theophylline and
theobromine.
The main advantages of these procedures are:
excellent chemical and isotopic purities, highly
reproducible yields (from labelled starting materials),
facile processing and purification, low cost and possi-
bility of labelling with other isotopes of hydrogen or
carbon.
1-Pivaloyloxymethyl-3-methylxanthine 23. Solid am-
monium formate (2.5 g, 39.7 mmol) was added in one
portion to a mixture of 1-pivaloyloxymethyl-7-benzyl-3-
methylxanthine 22 (2.5 g, 6.75 mmol) and palladium
on carbon (10%, 2 g) in dry methanol (100 ml) at 608C
under nitrogen. After 1 h heating under reflux, the
catalyst was removed by filtration, washed with hot
methanol (2 Â 100 ml) and the combined methanolic
solutions were evapored to give 23 (white solid, 1.63 g,
86%).
Acknowledgements
We would like to thank Murielle Jaubert and Jean
Guineton for GC/MS analysis, Dr Me´lanie Ethe`ve-
Quelquejeu for NMR spectroscopy and Dr Louis
Dehennin for helpful assistance in manuscript pre-
paration.
1-Pivaloyloxymethyl-3-7-dimethylxanthine-d3
24. A
mixture of 1-pivaloyloxymethyl-3-methylxanthine 23
(1.63 g, 5.81 mmol), potassium carbonate (1,7 g,
12.3 mmol) and CD3I (720 ml, 11.5 mmol) in dry DMF
(15 ml) was stirred for 2 h at 708C and then for 2 days at
room temperature in a sealed flask in the dark. The
solvent was evaporated, the remaining yellow oil was
dissolved in chloroform (100 ml) and filtered. The solid
was washed with chloroform (50 ml). After evaporation
down to 30 ml, the combined chloroform solutions were
REFERENCES
1. Houghton E. Biomed Mass Spectrom 1982; 9:
103–107.
2. Berlioz C, Falconnet JB, Desage M, Brazier JL. J
Label Compd Radiopharm 1987; 24: 275–289.
3. Kenani A, Bernier JL, Henichart JP. J Label Compd
Radiopharm 1995; 36: 187–192.
filtered through
a
silica gel pad and evaporated,
4. Pierce WM, Schlager JJ, Madden RJ, Hurst HE. J
Label Compd Radiopharm 1984; 21: 187–192.
5. Boukraa MS, Deruaz D, Bannier A, Desage M,
Brazier JL. J Label Compd Radiopharm 1995; 36:
773–787.
yielding a clear yellow oil. This oil was dissolved in
diethyl ether (50 ml), hexane (50 ml) was added and a
crystalline precipitate was formed. The volume was
reduced to 50 ml, and hexane (50 ml) was added. The
solution was evaporated down to 50 ml at room
temperature and stored at 48C for 12 h. White crystals
of 24 were collected by filtration, washed with cold
hexane and dried under vacuum at 908C (white
crystals, 1.51 g, 87%).
6. Falconnet JB, Brazier JL, Desage M. J Label Compd
Radiopharm 1986; 23: 267–276.
7. Cherrah Y, Falconnet JB, Desage M, Brazier JL,
Zini R, Tillement JP. Biomed Environ Mass Spec-
trom 1987; 14: 653–657.
8. Horning MG, Nowlin J, Thenot JP, Bouwsma OJ.
Stable Isotopes: Proceedings of the Third Interna-
tional Conference. Academic Press: New York,
1979; 379–384.
Theobromine-d3 17. 1-Pivaloyloxymethyl-3-7-dimethyl-
xanthine-d3 24 (1.4 g, 4.71 mmol) in 2.5 M sodium
hydroxide (50 ml) was heated under reflux for 2 h.
After cooling in an ice bath, the pH was adjusted to 5
with concentrated hydrochloric acid (precipitation
occurred). This suspension was stored at 48C for 12 h
and the precipitate of 17 was collected by filtration,
washed with cold water (2 Â 25 ml) and dried in vacuum
9. Tserng KY. J Pharm Sci 1983; 72: 526–529.
¨
10. Muller CE, Sandoval-Ramirez J. Synthesis 1995;
1295–1299.
11. Kramer GL, Garst JE, Mitchel SS, Wells JN.
Biochemistry 1977; 16: 3316–3321.
Copyright # 2007 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2007; 50: 33–41
DOI: 10.1002.jlcr

EASY PREPARATIVE SCALE SYNTHESES OF LABELLED XANTHINES 41
12. Miyamoto K, Yamamoto Y, Kurita M, Sakai R,
Konno K, Sanae F, Ohshima T, Takagi K,
Hasegawa T, Iwasaki N, Kakiuchi M, Kato I. J
Med Chem 1993; 36: 1380–1386.
26. Goldner H, Dietz G, Carstens E. Liebigs Ann Chem
1966; 69: 142–158.
27. Pfleiderer W, Kempter FE. Angew Chem Int Ed
1967; 6: 258–259.
13. Sakai R, Konno K, Yamamoto Y, Sanae F, Takagi K,
Hasegawa T, Iwasaki N, Kakiuchi M, Kato I,
28. Bridson PK. Synth Commun 1990; 20: 2459–2467.
29. Bridson PK, Wang X. Synthesis 1995; 855–858.
30. Kim HO, Ji X, Melman N, Olah ME, Stiles GL,
Jacobson KA. J Med Chem 1994; 37: 3373–3382.
31. Lespagnol A, Debaert M, Minard-Vaillant N. Chim
Miyamoto K.
4044.
J Med Chem 1992; 35: 4039–
14. Daly JW, Padgett W, Shamin MT, Butts-Lamb P,
Waters J. J Med Chem 1985; 28: 487–492.
15. Garst JE, Kramer GL, Wu YJ, Wells JN. J Med
Chem 1976; 19: 499–503.
´
Therap 1970; 321–326.
32. Papesch V, Schroeder EF. J Org Chem 1951; 16:
1879–1890.
¨
16. Lister JH. Austr J Chem 1979; 32: 387–397.
17. German Patent 352, 980 (1922 to E. Merck).
18. Agostini O, Bonacchi G, Coppini G, Di Marco G,
Paoli P, Toja E. Arzneim Forsch / Drug Res 1995;
45(I): 648–688.
33. Muller CE. Synthesis 1993; 125–128.
34. Pfleiderer W, Hutzenlaub W. Liebigs Ann Chem
1979; 1847–1854.
35. Traube W. Chem Ber 1900; 33: 3033–3052.
36. Black TH, Gatto C. Synth Commun 1989; 19:
843–850.
19. Hedayatullah M. J Heterocyclic Chem 1982; 19:
249–251.
37. Lagoja IM, Herdewijn P. Synthesis 2002; 301–314.
38. Fujii T, Saito T, Tamura K. Chem Pharm Bull 1991;
39: 2855–2862.
20. Gepner B, Kreps L. J Gen Chem (USSR) 1946; 16:
179–186.
21. Brederck H, Von Schuh HG, Martini AL. Chem Ber
1950; 83: 201–211.
39. Ram S, Spicer LD. Tetrahedron Lett 1987; 28:
515–516.
22. Ogilvie KK, Beaucage S, Gillen MF. Tetrahedron
Lett 1978; 35: 3203–3206.
40. Ueda T, Oda N, Sakakibara J, Takeya K. Hetero-
cycles 1982; 19: 2291–2294.
¨
23. Sidorov GV, Myasoedov NF. J Label Compd Radio-
pharm 1994; 34: 353–358.
41. Birkofer L, Ritter A, Kuhlthau HP. Chem Ber 1964;
97: 934–945.
¨
24. Bobranski B, Synowiedski Z. J Amer Pharm Assoc
Sci Ed 1948; 62–64.
42. Muller CE, Shi D, Manning Jr M, Daly JW. J Med
Chem 1993; 36: 3341–3349.
43. Organic Synthesis Coll. (1963); Vol. 4: 247; (1957);
37: 15.
25. Brederck H, Hennig I, Pfleiderer W, Weber G. Chem
Ber 1953; 86: 333–351.
Copyright # 2007 John Wiley & Sons, Ltd.
J Label Compd Radiopharm 2007; 50: 33–41
DOI: 10.1002.jlcr