A new method for synthesis of nolatrexed dihydrochloride

Article abstract of DOI:10.1021/op9002517

A new synthetic method for nolatrexed dihydrochloride (thymitaq) has been developed. The synthesis was accomplished in three steps featuring the direct conversion of the starting 4-bromo-5-methylisatin into the methyl anthranilate by potassium peroxydisul

Full text of DOI:10.1021/op9002517

Organic Process Research & Development 2010, 14, 346–350
A New Method for Synthesis of Nolatrexed Dihydrochloride
Xueqing Zhao,† Fei Li,‡ Weiping Zhuang,† Xiaowen Xue,*,§ Yuanyang Lian,† Jianhui Fan,† and Dongsheng Fang†
Fujian ProVincial Key Laboratory of Screening for NoVel Microbial Products, Fujian Institute of Microbiology,
Fuzhou 350007, P.R. China, Department of Medicinal Chemistry, Nanjing Medical UniVersity, Nanjing 210029, P.R. China,
and Department of Medicinal Chemistry, China Pharmaceutical UniVersity, Nanjing 210009, P.R. China
Abstract:
partially different route B: 2f3f6f5).8 The starting material
4-bromo-5-methylisatin (2) was prepared from 2-bromo-4-
nitrotoluene via the Sandmeyer reaction in the Webber method,
and now it is commercially available on the Chinese market
on 100 kilograms of scale. However, the reported preparation
of the free base 1 in the lab for drug screening did not provide
information about quality control, such as the identifications of
related substances, and the solutions to manufacturing problems.
Lately Chen and Wan presented three synthetic routes C, D,
and E on 10 grams scale (Scheme 1, the route C: 2f6f7f
the free base 1f1; the divergent route D: 2f3f8f7; the route
E: 2f3f8 (f7, one pot) f the free base 1), of which the
route E with one-pot synthesis from 8 to free base 1 was
believed the best.9 Unfortunately, there was not enough detailed
quality control information required for a product of pharma-
ceutical quality in this paper either.
A new synthetic method for nolatrexed dihydrochloride (thymitaq)
has been developed. The synthesis was accomplished in three steps
featuring the direct conversion of the starting 4-bromo-5-meth-
ylisatin into the methyl anthranilate by potassium peroxydisulfate/
sodium methoxide. In the final Ullmann reaction potassium
carbonate was employed in place of sodium hydride, and the
amount of copper catalysts was significantly reduced. Moreover,
sodium sulfide solution was utilized to efficiently remove copper
under approximately neutral conditions instead of hydrogen
sulfide/methanol under strongly acidic conditions. By means of
these modifications, nolatrexed dihydrochloride was ensured to
be prepared in good yield and high purity.
Recently Wennerberg and his co-workers reported a large-
scale synthetic process for 1 with a quality that meets the
specifications as pharmaceuticals (the route F: (2f)3f7f the
free base 1 f1).10 They explored a couple of reagents which
could be used to form quinazolinone 7 from the anthranilic acid
(3) and found 1-amidino-1,2,4-triazole hydrochloride was the
most convenient among the tested reagents. They also optimized
the Ullmann reaction by replacing NaH, which was assumed
to result in the formation of two impurities, with solid NaOH,
and employing excessive 2,4,6-trimercapto-2-triazine trisodium
salt (TMT-15) instead of H2S/methanol solution to get rid of
trace copper. Regardless of its many advantages, this elegant
approach suffered tedious isolation and purification procedures.
Herein we describe a concise synthetic method for 1 featuring
a direct conversion of 4-bromo-5-methylisatin into its methyl
anthranilate by potassium peroxydisulfate/sodium methoxide in
methanol. The anthranilate was then converted into quinazoli-
none 7 on kilogram scale through a well-established method
(Scheme 1, route A: 5f7). To ensure that nolatrexed was
obtained in high yield and purity, the amounts of copper
catalysts were minimized, and H2S was replaced with Na2S
solution as a copper scavenger for the final Ullmann reaction.
Introduction
2-Amino-6-methyl-5-(4-pyridylthio)-3H-quinazolin-4-one di-
hydrochloride (nolatrexed dihydrochloride, 1), which is also
known as thymitaq, is being developed for the treatment of
unresectable hepatocellular carcinoma (HCC).1 As a folate
analogue, thymitaq works as an inhibitor of thymidylate
synthase (TS) that directly binds the TS folate site, resulting in
the inhibition of DNA replication, DNA damage, S-phase cell
cycle arrest, and caspase-dependent apoptosis.2-4 Its phase II
clinical studies suggest that thymitaq has survival benefits in
patients with HCC.5,6 Currently thymitaq is under phase III
clinical trials.7
The first synthesis of nolatrexed (1, free base) was reported
by Webber at the early stage of drug discovery process (Scheme
1, the route A: 2f3f4f5f7·HClf the free base 1 and the
* Corresponding author. Telephone: (+86)-25-83199600. Fax: (+86)-25-
83199600. E-mail: xwenxue@cpu.edu.cn.
† Fujian Institute of Microbiology.
‡ Nanjing Medical University.
§ China Pharmaceutical University.
(1) Niculescu-Duvaz, L. Curr. Opin. InVest. Drugs 2001, 2 (5), 693–705.
(2) Chu, E.; Callender, M. A.; Farrell, M. P.; Schmitz, J. C. Cancer
Chemother. Pharmacol. 2003, 52 (suppl. 1), S80–89.
(3) Rustum, Y. M.; Harstrick, A.; Cao, S.; Vanhoefer, U.; Yin, M. B.;
Wilke, H.; Seeber, S. J. Clin. Oncol. 1997, 15, 389–400.
(4) Schiffer, C. A.; Clifton, I. J.; Davisson, V. J.; Santi, D. V.; Stroud,
R. M. Biochemistry 1995, 34, 16279–16287.
(8) (a) Webber, S. E.; Bleckman, T. M.; Attard, J.; Deal, J. G.;
Kathardekar, V.; Welsh, K. M.; Webber, S.; Janson, C. A.; Matthews,
D. A.; Smith, W. W.; Freer, S. T.; Jordan, S. R.; Bacquet, R. J.;
Howland, E. F.; Booth, C. L. J.; Ward, R. W.; Hermann, S. M.; White,
J.; Morse, C. A.; Hilliard, J. A.; Bartlett, C. A. J. Med. Chem. 1993,
36, 733–746. (b) Webber, S. E.; Bleckman, T. M.; Attard, J.; Jones,
T. R.; Varney, M. D. Antiproliferative Quinazolines. U.S. Patent
5,707,992, 1998.
(9) Chen, W.-M.; Wan, S.-H. Synth. Commun. 2007, 37, 53–61.
(10) Malmgren, H.; Ba¨ckstro¨m, B.; Sølver, E.; Wennerberg, J. Org. Process
Res. DeV. 2008, 12, 1195–1200.
(5) Staurt, K.; Tessitore, J.; Rudy, J.; Clendennin, N.; Johnston, A. Cancer
1999, 86, 410–414.
(6) Mok, T. S. K.; Leung, T. W. T.; Lee, S.-D.; Chao, Y.; Chan, A. T. C.;
Huang, A.; Lui, M.-C.; Yeo, W.; Chak, K.; Johnston, A.; Johnson, P.
Cancer Chemother. Pharmacol. 1999, 44, 307–311.
(7) Gish, R. G.; Wadler, S. J. Clin. Oncol. 2004, 22 (14S), 4260. ASCO
Annual Meeting Proceedings (Post-Meeting Edition), July 15, Supple-
ment.
346
•
Vol. 14, No. 2, 2010 / Organic Process Research & Development
10.1021/op9002517  2010 American Chemical Society
Published on Web 01/05/2010

Scheme 1. Synthetic routes A-F from 4-bromo-5-methylisatin (2) to nolatrexed dihydrochloride (1)
Results and Discussion
this idea in mind, we made a literature search, and found
a literature reference describing a simple direct conversion
from isatins to their anthranilates.13
By analyzing the structural characteristics of nolatrexed (1),
one can find that there usually are two critical steps to synthesize
1: (1) Ullmann reaction between the quinazolinone and 4-mer-
captopyridine; (2) the formation of the quinazolinone via the
cyclization of an appropriate intermediate, such as 3, 5, or 6
which may be derived from 4-bromo-5-methylisatin (2).
At first we took the formation of 2-aminoquinazolino-
nes into serious consideration. 2-Aminoquinazolinones can
be synthesized from anthranilic acids, anthranilates, isatoic
anhydrides, or anilines11 by reacting them with a number
of reagents such as cyanamide, haloformamidine and
guanidine. Among the reported methods, Chen’s synthetic
route E and the Wennerberg approach (the route F) were
believed best, but they suffered either complicated separa-
tion procedures, tedious purification, special reagents
applied, or low yields in the formation of the 2-amino-
quinazolinones (7 or 8) as discussed before. The routes
A and B (Scheme 1), however, gave us a piece of very
useful information that 7 · HCl was readily prepared in
high yield from 5 and chloroformamidine hydrochloride
(9), and the crude 7 · HCl could be easily isolated by
filtration and directly used for the next step without further
purification. Moreover, 9 could be prepared from very
cheap CaCN2.12 Naturally, we wondered whether there was
a simple method to transform 2 to 5 directly if the
oxidation of 2 was carried out in methanol rather than
water used in the conversation 2f3 in Scheme 1. Bearing
Next, we tried to employ this type of conversion into our
synthesis of 1 and designed a new route (G) from 2:
2f5f7·HClf the free base 1 f1 (Scheme 2). At the
beginning, the literature procedure from 2 to 5 was followed
with 30% H2O2 as an oxidizing agent, but the yield was not
satisfactory. Even if 98% H2O2 was employed, 5 was obtained
still in a moderate yield (60-70%). We found that in this
reaction the major side product was 3, which was postulated to
come from the hydrolysis of 5, as one molecule of H2O2 gave
off one molecule of water in the oxidation reaction. Therefore,
a peroxide oxidizing agent which does not generate water under
basic condition was desirable. It had been mentioned in the
patent that some peroxide compounds might be applicable to
this oxidation.13a K2S2O8 was then tried, instead of highly
explosive 98% H2O2, and a better yield (84.2%) and a good
purity of 5 (96.0%) were achieved. The major impurity (2.5%)
in 5 was identified as 6 by LC/MS in light of the integration to
the oxidation mechanism.13b
Compound 5 was then converted into 7·HCl according to
the literature procedure.8 Decomposition of 914 was not observed
in this case, and the yield did not decline when the reaction
scaled up to 1.2 kg of 5. The crude 7·HCl with 96.1% purity
(13) (a) Reissenweber, G.; Mangold, D. Preparation of alkyl anthranilate.
U.S. Patent 4,310,677, 1982. (b) Reissenweber, G.; Mangold, D.
Angew. Chem., Int. Ed. Engl. 1981, 20 (10), 882–883.
(14) 14. By following the reference,12 chloroformamidine hydrochloride
(9) was prepared in 84% yield; White crystals; Mp 134-138 °C dec;
IR (KBr, cm-1): 3463, 3033, 2248, 1692, 1612, 1413; CH4N2Cl2
(114.963) calculated: C 10.45, H 3.51, N 24.37; found C 10.27, H
3.59, N 24.65.
(11) Zeghida, W.; Debray, J.; Chierici, S.; Dumy, P.; Demeunynck, M. J.
Org. Chem. 2008, 73 (6), 2473–2475.
(12) Lecher, H. Z.; Kosloski, C. L. Halo-formamidine Salts and Method
of Preparation. U.S. Patent 2,727,922, 1955.
Vol. 14, No. 2, 2010 / Organic Process Research & Development
•
347

Scheme 2. Synthetic route G from 4-bromo-5-methylisatin (2) to nolatrexed dihydrochloride (1)
could be collected by filtration and directly used for the next
step without further purification.
9 with aqueous 28% NH3 solution to provide the free base
of 1 (nolatrexed) in 85% yield and 96.6% purity. The
content of the major impurity was 3.0%, which was
isolated from a couple of grams of the sample by flash
chromatography and primarily identified as 4,4′-dithio-
dipyridine.16 The content of heavy metals including copper
was less than 20 ppm. If the appearance of nolatrexed
looked gray or tan, or the content of heavy metals was
larger than 20 ppm, the above purification sequence would
be repeated. The main difference of this method from
others is the removal of copper under approximately
neutral conditions rather than strongly acidic conditions.
Under neutral conditions copper could be scavenged by
From our primary results of the Ullmann reaction
between 7 · HCl and 4-mercaptopyridine, we realized that
a large amount of Cu2O and CuBr might be mainly
responsible for the complicated isolation as well as for
the possible generation of more impurities. Intensive
investigation on Ullmann reactions revealed that cuprous
oxide or cuprous halides could be used in quantities as
low as 1-10% equiv, and mild bases, e.g. K2CO3 and
Cs2CO3, had been widely applied even when thiophenols
were used as reactants.15 What should be mentioned here
was that 4-mercaptopyridine was so expensive that the
preference was to utilize it as little as possible. Thus, we
decreased the amount of both Cu2O and CuBr to 2.5%
equiv, and 4-mercaptopyridine to 1.3 equiv. To avoid the
hydrolysis of the quinazolinone ring by caustic alkali,
anhydrous K2CO3 was employed instead of hazardous
NaH. The reaction was performed in DMA at 145 °C for
4 h. After the evaporation of the reaction mixture, the
resulted dark residue was dispersed in 95% ethanol and
adjusted to pH 4 with aqueous 36% HCl.
The removal of copper compounds from the Ullmann
reaction was usually carried out under strongly acidic
conditions when basic aqueous Na2S was employed as a
copper scavenger because 1 would precipitate as its free
base under basic conditions. However, we believed that
there would be a slight difference between the pH values
for copper (I, II) sulfide and 1 free base to precipitate
from Na2S solution. So we carefully adjusted the pH value
of the reaction mixture to 6-7 with saturated aqueous
Na2S and found dark copper (I, II) sulfide indeed
precipitated, whereas no white solid, i.e. nolatrexed,
precipitated. The dark mixture was filtered to give a
slightly yellow solution, which was further basified to pH
S
2- more completely than under strongly acidic conditions.
With the free base form in our hands, nolatrexed was
easily converted into its dihydrochloride form. To a
mixture of nolatrexed and ethanol was added an excess
of aqueous 36% HCl (2.5 equiv). The formed mixture was
then warmed to 50-60 °C to give a clear solution. Due
to the susceptibility of nolatrexed to acids, the solution
was filtered immediately and cooled down to room
temperature. Ethyl acetate was added to force more 1 to
precipitate. The mixture was kept at 0 °C overnight,
filtered to afford 1 in 90% yield with 99.7% purity. The
content of the only major impurity, which was unidenti-
fied, was 0.3%, and the copper content was less than 10
ppm. Fortunately, the major impurity detected in free base
1 had completely disappeared from 1 dihydrochloride.
Conclusion
We have developed a three-step synthetic method for
nolatrexed dihydrochloride (thymitaq) starting from 4-bro-
mo-5-methylisatin, with the formation of the methyl
anthranilate directly from 4-bromo-5-methylisatin by
potassium peroxydisulfate as the key step. In the Ullmann
reaction mild base anhydrous potassium carbonate was
employed instead of NaH, the amounts of cupric oxide
and cupric bromide were both reduced drastically, and
the copper compounds were removed more efficiently by
aqueous Na2S solution under approximately neutral condi-
(15) (a) Beletskaya, I. P.; Cheprakov, A. V. Coord. Chem. ReV. 2004, 238,
2237–2264. (b) Kim, J.-H.; Bang, S. W. Polym. J. 2000, 32 (2), 2000.
(c) Kametani, T.; Fukumoto, K. J. Chem. Soc. 1964, 6141–6146. (d)
Kuratsu, M.; Kozaki, M.; Okada, K. Chem. Lett. 2004, 33 (9), 1174–
1175. (e) McMorran, D. A.; Steel, P. J. Acta Crystallogr. 1998, C54,
1132–1133. (f) Jahng, Y.; Telford, J. R. Bull. Korean Chem. Soc. 2005,
26 (10), 1614–1615. (g) Sperotto, E.; Van Klink, G. P. M.; De Vries,
J. G.; Van Koten, G. J. Org. Chem. 2008, 73, 5625–5628. (h) Klapars,
A.; Buchwald, S. L. Org. Lett. 2001, 3, 3803–3805. (i) Kwong, F. Y.;
Klapars, A.; Buchwald, S. L. Org. Lett. 2002, 4, 581–584.
(16) For details about the formation of 4,4′-dithiodipyridine, see: Yama-
moto, T.; Sekine, Y. Can. J. Chem. 1984, 62, 1544–1547.
348
•
Vol. 14, No. 2, 2010 / Organic Process Research & Development

tions. Nolatrexed dihydrochloride of high purity and high
yield could be prepared via this concise synthetic method.
2-Amino-5-bromo-6-methyl-3H-quinazolin-4-one Hydro-
chloride (7 ·HCl). To a solution of 5 (1.20 kg, 4.92 mol)
in diglyme (10 L) was added 9 (566 g, 4.92 mol) at r.t.
with vigorous stirring. The mixture was heated at reflux
(160-165 °C) for 1 h and then cooled to 0 °C. The
precipitate was collected by filtration, washed with ether
(∼2 L), and finally dried in vacuum (35-40 mmHg) at
70 °C for 24 h to afford 1.24 kg of 7 · HCl as an off-
white powder; Yield: 86.7%; Purity: 96.1% (eluents:
CH3CN-H2O ) 10-90, pH 4.94; Rt ) 37.9 min); Rf )
0.59 [ethyl acetate/(0.63 M NH3 in ethanol) ) 6/4]; Mp
> 300 °C (lit.8a Mp > 390 °C and yield: 88%); IR (KBr
cm-1): V˜ 3170, 1716, 1658, 1470, 812; 1H NMR (DMSO-
d6): δ 2.40 (s, 3 H, -CH3), 7.06 (s, 1 H, N-H), 7.33 (d,
J ) 8.6 Hz, 1 H, Ar-H), 7.69 (d, J ) 8.6 Hz, 1 H, Ar-H),
Experimental Section
Melting points were determined with a Bu¨chi 540
melting point apparatus and are uncorrected. TLC was
performed on glass plates (GF254, 50 mm × 100 mm). IR
spectra were run on FI-IR spectrometer (Perkin-Elmer)
with oily samples as film and with solid samples as pellets
1
(KBr). H NMR spectra were recorded on Bruker AV-
400. LC/MS analysis for methyl 2-amino-6-bromo-5-
methylbenzoate (5) was performed on LC/MS Agilent
1100 series charged with 4.6 mm × 250 mm (5 µ) and
precolumn 4.6 mm × 25 mm. HPLC analyses for the
2-aminoquinazolinone hydrochloride (7 · HCl), nolatrexed
(1, free base), and nolatrexed dihydrochloride (1) were
executed on Shimadzu CLASS-VP charged with Nucleosil
C18 (4.6 mm × 250 mm, 7 µ) at 254 mn of detection
wavelength and at 1 mL of flow rate. The gradient elution
for nolatrexed dihydrochloride (1) was employed, ramping
from 88/12 [phosphate buffer (pH 4.94, 100 mg of
Na2HPO4 and 117.5 g of NaH2PO4 in 1000 mL of water)/
acetonitrile] to 72/28 during 35 min. Each injection
volume was 10 µL at 1 mg/mL of concentration. Silica
GF254 for TLC was purchased from Qingdao Marine
Chemical Company, China.
1
8.00 (br s, 2 H, NH2), 12.0 (br s, 1 H); HNMR (DMSO-
d6 -D2O): δ 2.34 (s, 3 H, -CH3), 7.18 (d, J ) 8.4 Hz, 1
H, Ar-H), 7.61 (d, J ) 8.4 Hz, 1 H, Ar-H); MS (ESI+)
m/z: 256.0, 254.0 [M - Cl]+; (ESI-) m/z: 253.9, 251.9
[M - Cl - 2]+; The content of one unknown major
impurity was 3.2%.
2-Amino-6-methyl-5-(4-pyridylthio)-3H-quinazolin-4-
one (Nolatrexed; the Free Base 1). To a mixture of
4-mercaptopyridine (396 g, 3.56 mol), Cu2O (0.99 g, 6.9
mmol), and CuBr (0.98 g, 6.9 mol) in anhydrous DMA
(8 L) were added anhydrous K2CO3 (1.14 kg, 8.25 mol)
and 7 · HCl (800 g, 2.75 mol). The mixture was heated at
140-145 °C for 4 h, and most of the solvent was then
removed by distillation under reduced pressure. A brown
residue was dispersed with 95% ethanol (4 L) and H2O
(4 L) while warm, and acidified with aqueous 36% HCl
solution (∼2 L) to pH 4 when cooled below 20 °C. To
the acidic mixture was added dropwise aqueous saturated
Na2S solution (680 mL) to reach pH 6-7 to remove copper
compounds until a white precipitate (nolatrexed) emerged.
Charcoal (40 g) was added, and stirring was continued
for 15 min. A slightly yellow solution was obtained by
filtration and basified to pH 9 with aqueous 28% NH3
solution (1.60 L). The precipitate was collected by
filtration, washed with water, and dried under vacuum
(35-40 mmHg) at 100-110 °C for 24 h to yield 665 g
of nolatrexed (1, free base) as a white powder; Yield:
85.0%; Purity: 96.6% (eluents: CH3CN-H2O ) 10-90,
pH 4.94; Rt ) 11.8 min); Rf ) 0.31 [ethyl acetate/(0.63
M NH3 in ethanol) ) 6/4]; Mp 300-302 °C (lit.:8a a tan
solid; Mp 301-302 °C); MS (ESI+) m/z: 285.1 [M + 1]+;
the major impurity: 3.0% (Rt ) 13.0 min); Mp 73-77
4-Bromo-5-methylisatin was purchased from Medicalchem
Co. Ltd. (Yencheng, China). N,N-Dimethyl acetamide (DMA)
and diglyme were dried over 4 Å molecular sieves for more
than 10 days. All reactions were carried out under nitrogen
protection.
Methyl 2-Amino-6-bromo-5-methylbenzoate (5). To a
mixture of 2 (1.60 kg, 6.67 mol) and anhydrous methanol
(6.70 L) was added sodium methoxide (22.6%, 4.80 kg,
20.1 mol) to give a violet solution, to which K2S2O8 (1.90
kg, 7.03 mol) was added in parts below 10 °C with an
ice-water bath. After addition the reactant mixture turned
yellow, and stirring was continued for 1 h at r.t. The
reaction mixture was then adjusted to pH 8-9 with
aqueous 36% HCl (1.24 L) below 15 °C and the excessive
K2S2O8 was destroyed by aqueous 5% Na2S2O4 solution
(450 mL). After rotary evaporation under a reduced
pressure at 55 °C, a brown liquid was left, and it was
mixed with CH2Cl2 (6 L) and H2O (4 L). The organic
phase was separated, and the aqueous phase was extracted
with CH2Cl2 (4 L). The combined organic phases were
dried over Na2SO4 and concentrated with a rotavapor. The
resulted brown liquid was further distilled under high
vacuum to give 1.37 kg of 5 as a colorless liquid
(165-175 °C/6-8 mmHg); Yield: 84.2%; Purity: 96.0%
(Rt ) 4.21 min); MS (ESI+) m/z: 246.0, 244.0 [M + 1]+;
IR (KBr cm-1): V˜ 3472, 3382, 2953, 2924, 1716, 1622,
1480, 1277, 816; 1H NMR (CDCl3): δ 2.28 (s, 3 H,
Ar-CH3), 3.91 (s, 3 H, -OCH3), 4.26 (br s, 2 H, NH2),
6.54 (d, J ) 8.2 Hz, 1 H, Ar-H), 7.00 (d, J ) 8.2 Hz, 1
H, Ar-H). One major impurity: 2.48% (Rt ) 5.24 min);
MS (ESI+) m/z: 260, 258 [M + 1]+.
1
°C; H NMR (DMSO-d6): δ 7.95 (d, J ) 6.4 Hz, 4 H),
8.81 (d, J ) 6.4 Hz, 4 H); MS (ESI+) m/z: 219.2 [M -
1
1]+; The H NMR data were identical to those for 4,4′-
dithiodipyridine in the database,17 and this impurity was
determined to be 4,4′-dithiodipyridine on the basis of its
1
mp, MS, and H NMR data.
2-Amino-6-methyl-5-(4-pyridylthio)-3H-quinazolin-4-
one Dihydrochloride (1, Nolatrexed Dihydrochloride). To
a mixture of nolatrexed (500 g, 1.76 mol) in ethanol (6
L) was added 36% HCl (378 mL, 4.40 mol). The mixture
was warmed to a clear solution to 50-60 °C and then
Vol. 14, No. 2, 2010 / Organic Process Research & Development
•
349

filtered immediately when warm. The mixture was cooled
down to r.t., and ethyl acetate (6 L) was added. The
mixture was kept at 0 °C overnight. The precipitate was
collected by filtration, washed with cooled anhydrous
ethanol, and finally dried under vacuum (35-40 mmHg)
at 60 °C for 12 h to provide 567 g of 1 as white crystals;
Yield: 90.2%; Purity: 99.7% (Rt ) 8.72 min); IR (KBr
Acknowledgment
We thank the Major Project Founds of Fujian Province
(2004Y002) and China Pharmaceutical University (211080) for
the support of this work.
Supporting Information Available
Spectroscopic data for compounds 1, 5, 7, and 9 and for
impurity 4,4′-dithiodipyridine. This material is available free
of charge via the Internet at http://pubs.acs.org.
1
cm-1): V˜ 3401, 3058, 2929, 1701, 1621, 1471, 799; H
NMR (DMSO-d6): δ 2.43 (s, 3H, -CH3), 7.53 (d, J )
6.9 Hz, 2H, Pyr-H), 7.67 (d, J ) 8.5 Hz, 1H, Ar-H),
7.92 (d, J ) 8.5 Hz, 1 Hz, Ar-H), 8.30 (br s, 3H, NH3),
8.52 (d, J ) 6.9 Hz, 2H, Pyr-H); MS (ESI+) m/z: 285 [M
- 1-2Cl]+; (ESI+) m/z: 283 [M - 1- 2HCl]+. See
Supporting Information for its TGA and DSC. The content
of one unknown major impurity was 0.30% (Rt ) 17.0
min).
Received for review September 24, 2009.
OP9002517
(17) Spectrum ID: WHSP48928 in Integration Spectral Database System
of Organic Compounds. Data were obtained from the National Institute
of Advanced Industrial Science and Technology.
350
•
Vol. 14, No. 2, 2010 / Organic Process Research & Development