Methanol (Caledon), acetone (Caledon), acetaldehyde (BDH), aceto-
nitrile (Caledon), phosphoric acid (BDH, 85%), starch (Fisher, soluble
potato), phthalazone (Fluka), phthalazine (Aldrich), D-galactose (Sigma),
and D-glucose (Aldrich) were used as supplied. Hydralazine hydro-
chloride and cornstarch were obtained from suppliers and were used
as supplied.
Solid -Sta te F lu or escen ce An a lysis of Aged Ta bletssTen aged
tablets were finely powdered and placed in a fritted glass funnel
(medium pore size). The powder was washed using methanol and
sonication. A portion of placebo powder was also washed with
methanol. They were then mixed with water (4:1, w/w), respectively,
for the solid-state fluorescence study.
selective 1H decoupling showed that the carbon was directly bonded
to the proton at 11.527 ppm), 135.8 (CH), 133.9 (CH), 130.2 (CH),
124.1 (C), 123.1 (CH), 120.0 (C), 39.8 (CH3); IR (KBr) 1627.6, 1601.5
cm-1; MS (EI) m/z 184 (43), 157 (13), 129 (70), 115 (100), 101 (27), 88
(26).
P r ep a r a t ion of 3-Met h yl-s-t r ia zolo[3,4-a ]p h t h a la zin e (V)s
Sodium acetate (0.8 g) and I (1 g) in 30 mL of water were mixed with
acetaldehyde (0.29 g) and refluxed in water for 20 min. The yellow
solution formed was extracted with dichloromethane (2 × 15 mL).
Evaporation of the dried extracts (anhydrous sodium sulphate),
followed by recrystallization from isopropyl alcohol, afforded V (0.3
g) as yellowish crystals: mp 168-170 °C; IR (Nujol) 1626 cm-1
.
Compound V had been prepared earlier by reacting hydralazine and
Rea ction of I w ith Sta r ch in Wa ter sStarch (5 g) and water (200
mL) in a 200 mL round-bottomed flask were heated at 95 °C for 5
min. To the opaque fluid was added hydralazine hydrochloride (I)
(0.5 g). The mixture was magnetically stirred for 2 days at room
temperature and then concentrated at 60 °C under reduced pressure
until about 50 mL of water remained. Methanol (150 mL) was added
and the resulting mixture stirred vigorously for 30 min. After the
aqueous methanolic solution, containing unreacted I, was decanted,
the residue was evaporated to dryness at 60 °C under reduced
pressure and ground to a powder. The powder was transferred to a
fritted glass funnel (medium pore size) and mixed with 5 mL of
methanol. The mixture was sonicated until all the methanol passed
through the funnel. More methanol was added to the funnel and the
procedure was repeated until the methanol wash contained no
hydralazine hydrochloride as determined by HPLC analysis. The
dried powder was thoroughly mixed with water (4:1, w/w) for the solid-
state fluorescence study. (The water was added to minimize any
background signals due to starch).
R ea ct ion of I w it h St a r ch in Dim et h ylsu lfoxid e a n d GP C
An a lysissStarch (5 g) in 200 mL of anhydrous DMSO was stirred
magnetically at room temperature for 36 h. The dispersion was
filtered and a portion of the filtrate (150 mL) was mixed with I (0.05
g). After the mixture was sonicated for 10 min the resultant yellowish
solution was heated for 1 h and kept at room temperature for one
day. The solution showed fluorescence emission at about 417 nm
when it was excited at 320 nm. A portion of the solution was diluted
with THF (at a 1:1 ratio) and analyzed with GPC. A fraction with tR
) 20 min was detected in addition to a signal with tR of 25.5 min.
Under the same chromatographic conditions the retention time of
triazolophthalazine (IIIb) and 3-(D-galacto-pentitol-1-yl)-1,2,4-tri-
azolo[3,4-a]phthalazine [IIIa , R ) -(CHOH)4CH2OH] was 25.5 min.
ethylacetate,5 mp 171-172 °C.
Aceton e P h th a la zin -1-ylh yd r a zon e (VII)sI (1 g) in water (30
mL) was mixed with acetone (0.38 g) at room temperature and the
solution was magnetically stirred for 10 min. VII which precipitated
was collected by filtration as yellow crystals (0.54 g, dried over
CaCl2): mp 116 °C; lit.5 mp 114 °C; IR(Nujol) 1633.8, 1614.5 cm-1
.
HP LC Qu an titation of Hydr alazin e-Related Com pou n dssThe
HPLC system consisted of a Waters 510 pump and a Waters 481
variable wavelength detector set at 220 nm. The analytical column
used was a Bondapak C18 (300 mm × 3.9 mm i.d., 10 m particle
size), with a mobile phase made up of CH3CN/C8H17SO3Na (5 mM)/
H3PO4 (150:850:0.45, v/v/v). Chromatography was performed at room
temperature using a flow rate of 2.0 mL/min.
Quantitation of Related Compounds in TabletssSix aliquots of the
related compounds standard solution (0.5 g/mL each of phthalazone,
phthalazine, hydralazine hydrochloride, and IIIb) were injected.
After the system was deemed to be suitable for use, the related
compounds test solution, which was prepared by extracting the tablet
powder with a mixture of CH3CN/5 mM C8H17SO3Na (150:850, v/v)
in a volumetric flask with sonication, was injected into the chromato-
graph. The amounts of phthalazone, phthalazine, and IIIb in the
test solution were calculated by an external standard method. The
amounts of other unknown impurities were expressed in terms of
hydralazine assuming they have the same molar extinction coef-
ficients at 220 nm.
Quantitation of IIIc in Aged Tablets in Terms of IIIbsA portion
of fine powder of 20 aged hydralazine hydrochloride tablets, equivalent
to 50 mg of hydralazine hydrochloride, was mixed with acetone (15
mL) and vigorously stirred at room temperature for 14 h to quench
the remaining hydralazine hydrochloride in the tablet powder. The
solid residue which contained IIIc was separated, by filtration, from
the yellow acetone solution which contained acetone phthalazin-1-
ylhydrazone (VII). The residue was washed with acetone thoroughly
(5 × 15 mL) and dried under vacuum. It was mixed with 60 mL of
0.01 N hydrochloric acid and refluxed for 13 h. The resultant aqueous
mixture was neutralized with 1 N NaOH and diluted to 250 mL with
acetonitrile. An aliquot of the solution was centrifuged (rpm ) 2000)
and 1 mL of the supernatant solution was diluted 25-50 times. An
aliquot of the final solution was injected into the chromatograph.
3-(D-ga la cto-P en titol-1-yl)-1,2,4-tr iazolo[3,4-a ]ph th alazin e [IIIa,
R ) -(CHOH)4CH2OH)]sIIIa was prepared by heating D-galactose
and I in methanol/water for 20 min as reported3 and obtained as
yellowish crystals: IR (Nujol) 1632 cm-1; lit.3 1630 cm-1. An aqueous
solution of this compound showed a fluorescent emission at 415 nm.
Isola tion of s-Tr ia zolo[3,4-a ]p h th a la zin e (IIIb), a Degr a d a n t
of IIIa sA yellowish fluorescent solution (
>400 nm) was obtained
em
by heating I (10 mg) and D-glucose (10 mg) in water (10 mL) for 10
min, indicating formation of 3-(D-gluco-pentitol-1-yl)-1,2,4-triazolo-
[3,4-a]phthalazine [IIIa , R ) -(CHOH)4CH2OH)]. In a separate
experiment, I (1 g), D-glucose (100 g), and water (5 g) were heated at
90 °C for 24 h. The reaction mixture was diluted with water and
extracted with chloroform. The organic layer was washed with water,
dried, and evaporated under reduced pressure. Recrystallization of
the solid residue from isopropyl alcohol (three times) afforded IIIb
as colorless crystals: mp 189 °C; 1H NMR (CDCl3) 9.033 (s, 1H),
8.731 (d, 1H, J ) 7.9 Hz), 8.650 (s, 1H), 7.977 (m, 1H), 7.942 (d, 1H,
J ) 7.8 Hz), 7.839 (m, 1H); 13C NMR (CDCl3) 148.1 (CH), 142.5 (C),
139.7 (CH), 134.2 (CH), 131.0 (CH), 128.1 (CH), 123.5 (C), 123.4 (CH),
123.3 (C); MS (EI) m/z 170 (M+, 50), 115 (M+ - CHN3, 100), 88 (35);
IR (KBr) 1627.5 cm-1; high resolution MS m/z calcd for C9H6N4,
170.0594; found, 170.0597. Compound IIIb had been prepared earlier
by reacting hydralazine and formaldehyde, mp 190-191 °C.5
Results and Discussion
When a thick aqueous paste of expired tablets was excited
at 320 nm an extremely weak fluorescence emission at about
414 nm was observed (see inset, Figure 1). To determine
whether hydralazine hydrochloride and starch would lead to
fluorescing compounds, these two entities were mixed with
water and stored for 2 days after which the water was
evaporated. The severe conditions, although unrealistic with
respect to solid state (tablets) conditions, would produce IIc
and/or IIIc in a timely manner. Such a product, strictly
speaking, a mixture containing starch, degradants of starch,
IIc, IIIc, etc., was needed in the early stage of our investiga-
tion to find suitable excitation and detection parameters.
When the solid residue, which had been thoroughly washed
with methanol to remove any remaining hydralazine hydro-
chloride, was excited at 320 nm, fluorescent emission at 414
nm was detected (Figure 1a). A thick aqueous paste of starch,
which was prepared and treated similarly except that no
hydralazine hydrochloride was used, showed virtually no
F or m a tion of s-Tr ia zolo[3,4-a ]p h th a la zin e (IIIb) via Decom -
p osition of IIIcsHeating I (1 g), corn starch (100 g), and water (5
g) at 95 °C for 24 h also produced IIIb, as determined by HPLC
analysis. Chromatography of the reaction mixture afforded an
unknown compound VI: 1H NMR (DMSO-d6) 11.527 (s, 1H, this
proton underwent exchange in the presence of D2O), 9.571 (s, 1H),
8.579 (d, 1H, J ) 7.9 Hz), 8.424 (d, 1H, J ) 7.9 Hz), 8.237 (m, 1H),
8.182 (m, 1H), 4.389 (s, 3H, saturation of the CH3 resonance during
an NOE experiment resulted in an enhancement of the proton at
11.527 ppm); 13C NMR (DMSO-d6) 153.1 (CH), 142.3 (C), 139.8 (CH,
Journal of Pharmaceutical Sciences / 327
Vol. 85, No. 3, March 1996