Professional Documents
Culture Documents
doi: 10.1093/chromsci/bmx036
Article
Article
Abstract
Unknown impurities were identied in ibuprofen (IBU) soft gelatin capsules (SGCs) during long-
term stability testing by a UHPLC method with UV detection and its chemical formula was deter-
mined using high resolution/accurate mass (HRAM) LCMS. Reference standards of the impurities
were subsequently synthesized, isolated by semi-preparative HPLC and characterized using HRAM
LCMS, NMR and IR. Two impurities were formed by esterication of IBU with polyethylene glycol
(PEG), which is used as a ll of the SGCs, and were identied as IBUPEG monoester and IBUPEG
diester. Two other degradants arised from reaction of IBU with sorbitol and sorbitan, which are
components of the shell and serves as plasticizers. Thus, IBU sorbitol monoester (IBUsorbitol) and
IBU sorbitan monoester (IBUsorbitan ester) were identied. An UHPLC method was further opti-
mized in order to separate, selectively detect and quantify the degradation products in IBU SGCs.
The Author 2017. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com 1
2 Doua et al.
Several HPLC methods for IBU determination and its related com- Experimental
pounds and impurities in pharmaceutical preparations have been
described in the literature. The most common technique for IBU
Reagents and chemicals
quantication is RP-HPLC using an octadecylsilane column as sta- Acetonitrile HPLC gradient grade (J.T. Baker, USA), methanol (J.T.
tionary phase and UV detection (1316). Capillary electrochromato- Baker, USA) and water puried by Milli-Q system (Millipore, USA)
graphy using RP-18 packed capillary was used for simultaneous were used for preparation of samples, reference solutions and
separation of IBU and its impurities (Ph. Eur. impurity A, B, C, E) as mobile phases. All other chemicals (ortho-phosphoric acid, formic
well (17, 18). acid, dichlormethane, dicyclohexylcarbodiimide and 4-dimethylami-
Although degradation chemistry of IBU API and IBU in tablet nopyridine) were of analytical grade or pure grade quality (Sigma-
formulations was well described (1822), only limited data related Aldrich, Czech Republic). The standard of IBU (in-house standard,
to stability of IBU in SGCs formulation has been reported (23). purity 99.8%) was prepared in Zentiva (Czech Republic). PEG 600
Forced degradation of IBU was observed in tablets containing PEG (range of average molecular weight: 570630, dynamic viscosity at
and polysorbate 80. Unfortunately, the structures of IBU degrada- 20C of 50% solution: 18 mPa s, hydroxyl value: 188 mg KOH/g)
tion products were not determined. Moreover, only accelerated deg- was obtained from Clariant Produkte (Germany). Polysorb 85 con-
radation but not long-term stability data were reported (24). tained 43% of D-sorbitol and 27% of 1,4-sorbitan and was obtained
Decrease of esterication reaction between IBU and PEG was from Roquette (France). The shell of in-house prepared SGCs com-
described when PEG in combination with polyvinylpyrrolidone was posed of gelatin, partially dehydrated sorbitol liquid, carmine red
used as a SGCs ll (25). Formation of IBUPEG ester and its hydro- and puried water in mass ratio 44:22:0.05:34. The ll of the SGCs
lysis was described in EMEA withdrawal assessment report for com- composed of IBU, PEG 600, potassium hydroxide 85% and water
bination product IBU/diphenhydramine hydrochloride (26). Besides in mass ratio 58:29:8:5.
that, IBUPEG esters were prepared and their potential as a drug
prolonged release system was investigated. However, these studies
HPLC instrumentation and methods
were aimed solely to preparation of new IBU polymeric prodrugs
All chromatographic experiments were carried out on an Acquity
and studying of the IBU degradation was not the concern (27, 28).
UPLC system with a photodiode array detector (Waters, USA). The
Further, the various types of poly(ethylene glycol) (PEG)IBU conju-
system control, data acquisition and processing was accomplished
gates by the nucleophilic substitution of bromo-terminated PEG
by the Empower software (Waters, USA). Chromatographic separa-
with IBU-Cs salt were synthesized (29).
tions were performed on an Acquity BEH C18 column (100
This article describes identication of degradation products
2.1 mm2, 1.7 m; Waters, Czech Republic). The mobile phase was
which were found in IBU SGCs pharmaceutical preparations
pumped at 0.5 ml/min, the injection volume was 1 l and analytes
using UHPLC method with UV detection. The impurities were
were monitored at a wavelength of 220 nm. The gradient elution
synthesized, puried by semi-preparative HPLC and character-
employed solutions A and B as mobile phase components. Method
ized using HRAM LCMS, NMR and IR analysis. All character-
I: The solvent A was 0.1% ortho-phosphoric acid, while solvent B
ized degradants are esters formed by drug-excipient interactions.
Two degradants were formed by esterication of IBU with ll of
the SGCs and were identied as IBUPEG monoester and IBU
PEG diester. Two other impurities arised from reaction of IBU
with content of the shell and to best of our knowledge has not
been reported yet. These unknown degradants were determined
as IBU sorbitol monoester (IBUsorbitol) and IBU sorbitan mono-
ester (IBUsorbitan).
Figure 3. Preparation of IBUPEG esters (A), IBUsorbitol (B) and IBUsorbitan (C).
was acetonitrile. The gradient program was set as follows: time/% Jose, USA) coupled to an HPLC HTS PAL system (CTC Analytics,
of solvent B: 0/10, 2.0/10, 17/80, 20/80, 20.5/10 with an equilibra- Switzerland). LC separation was performed on a Kinetex C18, 150
tion time of 2 min. The column was thermostated at a temperature 4.6 mm2, 5 m (Phenomenex, Torrance, USA) column using 1.0 ml/min
of 35C. Method II: The solvent A was 0.1% ortho-phosphoric acid ow rate. The gradient elution employed solutions A and B as mobile
and solvent B was acetonitrile. The gradient program was set as fol- phase components. The solvent A was 0.1% formic acid, while sol-
lows: time/% of solvent B: 0/5, 60/90, 61/5 with an equilibration vent B was acetonitrile. The gradient program was set as follows:
time of 2 min. The column was thermostated at a temperature of time/% of solvent B: 0/43, 17/43, 25/100, 30/100, with an equilibra-
45C. Method III: The solvent A was water and solvent B was meth- tion time of 7 min. For an ionization of the analytes an APCI ion
anol. The gradient program was set as follows: time/% of solvent B: source was operated in the positive ion mode (desolvation tempera-
0/0, 70/85, 90/85, 91/0 with an equilibration time of 4 min. The col- ture 400C, capillary temperature 300C, discharge current 4 A and
umn was thermostated at a temperature of 65C. tube lens voltage 40 V).
4
Position (C) (H) Mult. Integ. H J(H,H) Position (C) (H) Mult. Integ. H J(H,H) (C) (H) Mult. Integ. H J(H,H)
Note: in ppm, J in Hz. Reference for IBUPEG monoester 1H and 13C was CDCl3 signal. Reference for IBUsorbitol and IBUsorbitan was DMSO signal: 1H (DMSO) = 2.5 ppm and 13C (DMSO) = 39.50 ppm.
Doua et al.
Esterication of Ibuprofen in Soft Gelatin Capsules FormulationsIdentication, Synthesis and Liquid 5
Impurity Measured Theoretical Mass error Molecular formula [M + H]+ Fragments/adducts Fragments/adducts formulas
mass (m/z) mass (m/z) (ppm) mass (m/z)
Impurity IR
IBUPEG monoester (OH) 3500, (CH) 2953, 2867, (C = O) 1732, (C = C) 1512, 1456, (COC) 1201, (CO) 1095
IBUsorbitol (OH) 3360, (CH) 2951, 2867, (C = O) 1702, (C = C) 1512, 1462, (COC) 1213, (CO) 1089, 1024
IBUsorbitan (OH) 3402, (CH) 2952, 2918, 2865, (C = O) 1723, 1700, (C = C) 1512, 1464, (COC) 1168, (CO) 1064, 1043
2,545 binary gradient module, 2,767 sample manager, 515 HPLC Results
pump, MS and 2,487 dual-wavelength absorbance detector; Waters,
USA). The ow rate of 20 ml/min was employed throughout the
Identication of unknown degradation products
preparation. The injection volume was 900 l. In order to monitor Unknown degradation products were detected during the long-term
UV signal from the semi-preparative column, the efuent was stability testing of the IBU SGCs formulation in climatic chamber at
splitted in the ratio 1:1,000 into the methanol ow from 515 HPLC 25C and 60% relative humidity (Fig. 1). Gradient UHPLC method
pump, which was directed to MS and UV detector. A semi- (Method I) was used showing relative retention time (RRT) 0.78,
preparative XBridge Prep C18 OBD column (100 19 mm2, 5 m; 0.86, 1.05, 1.521.63 in respect to IBU. The similar impurities pro-
Waters, USA) was used for preparative purposes. le was observed for IBU SGCs developed by Zentiva and obtained
Method IV was used for isolation of IBUPEG monoester: from original manufacturer (data not shown).
The fraction collection was triggered by setting a minimal inten- Subsequently, HRAM LCMS measurement was performed in
sity threshold (MIT = 3,500 V for UV detection) of the UV sig- order to investigate chemical structure of the unknown degradants.
nal at 265 nm. The mobile phase consisted of 0.025% aqueous HPLC peak corresponding to impurity RRT 1.05 was identied
solution of ammonium hydroxide (solvent A) and acetonitrile as ester of IBU and PEG (IBUPEG monoester). PEG is used as a ll
(solvent B). The separation was performed using gradient elution of the SGCs and the ratio of IBU with PEG in the formulation is
based on following program: time/% of solvent B: 0/50, 2.0/50, 2:1. The polymeric compound consisted of a varying number of
4.4/80, 4.6/100, 6.5/100, 7.0/50 with a re-equilibration time of repeating ethoxy units between terminal hydroxyl group and IBU
1.0 min. Method V was used for preparation of IBUsorbitol and group (Fig. 2A). The isotopic pattern was represented by typical
IBUsorbitan: The fraction collection was triggered by setting a polymeric envelope with distance corresponding to ethylenglycol
minimal intensity threshold (MIT = 15,000 V for UV detection) monomer unit (44, C2H4O). Molecular formula was identied as
of the UV signal at 265 nm. The mobile phase consisted of 0.1% C2n+13H4n+18On+2 and monoisotopic mass was ranging from
aqueous solution of formic acid (solvent A) and acetonitrile (sol- 514.3142 (for n = 7) to 1,174.7074 (for n = 22). The range of
vent B). The separation was performed using gradient elution molecular weight was in agreement with specication of PEG 600,
based on following program: time/% of solvent B: 0/35, 0.5/35, which was used in the formulation.
7.5/42, 7.7/100, 8.7/100, 8.9/35 with a re-equilibration time of Moreover, degradation product RRT 1.521.63 was identied
1.1 min. as conjugate of two IBU molecules connected by PEG via ester
bonds (IBUPEG diester, Fig. 2B). Molecular formula was deter-
mined as C2n+26H4n+34On+3 and monoisotopic mass was ranging
NMR instrumentation and methods from 702.4343 (for n = 7) to 1,362.8275 (for n = 22). The isotopic
Nuclear magnetic resonance (NMR) spectra were obtained using a pattern was again represented by typical polymeric envelope.
Bruker Avance 500 (Bruker Biospin, Germany) at 500.13 MHz Two other degradants arised from reaction of IBU with sorbi-
(1H) and 125.78 MHz (13C) with Prodigy probe in CDCl3 at tol and sorbitan, which are components of the shell and serves as
298 K. plasticizers. Impurity RRT 0.78 was identied as IBU sorbitol
6 Doua et al.
Preparation of IBUsorbitan ester portions. After all CDI had been added, the solution was stirred for
Polysorb 80 (85/70/00) was concentrated in rotary evaporator under 10 min and then slowly added to the solution of polysorb in DMSO.
reduced pressure at 50C to obtain colorless honey, in which 6.6% Resulting clear slightly yellow solution was stirred for 60 h. This
of water was found using Karl Fischer titration method. This con- mixture was puried using semi-preparative HPLC. Collected frac-
centrated polysorb was used for synthesis. A 8.94 g of preconcen- tions containing desired product were evaporated at 40C under
trated polysorb 80 was dissolved in 5 ml warm DMSO, and then reduced pressure. Resulting oil was dried in vacuum of oil pump.
4.58 g (28 mmol) carbonyldiimidazole (CDI) was added to the Yield was 0.4 g of IBUsorbitan ester.
warm solution in portions. Mixture was foaming due to evolution Methods of preparation of IBUPEG, IBUsorbitol and IBUsor-
of carbon dioxide. To the separately prepared solution of 5.16 g bitan esters were not further optimized due to preparation of analyt-
(25 mmol) IBU in 3 ml DMSO was added 4.05 g (25 mmol) CDI in ical standards only.
Esterication of Ibuprofen in Soft Gelatin Capsules FormulationsIdentication, Synthesis and Liquid 7
The characterization of reference standards modied by changing the organic solvent (methanol), gradient prole
In order to prepare in-house standards of IBUPEG monoester, and increasing the temperature to achieve baseline resolution of the
IBUsorbitol ester and IBUsorbitan ester, the structures of prepared individual oligomers. UHPLC methods were used for the chemical
impurities were conrmed and characterized by NMR, MS and IR characterization of IBUPEG monoesters (Method III) and IBUPEG
analysis (Tables IIII). The assignment of NMR signals was per- diesters (Method II). The separation of the corresponding IBUPEG
formed by means of 1H, 13C and 2D (COSY, HSQC and HMBC) esters is shown in Figure 4, including the evaluation of percentage of
spectra. NMR data unambiguously conrmed the proposed struc- individual oligomers in IBUPEG esters by internal normalization.
tures of IBU esters. According to the intensities of the NMR signals, The Method II was successfully applied for determination of related
average number n of PEG units in IBUPEG monoester was 13. substances in pharmaceutical SGCs formulations.
Moreover, the elucidation of chemical structure by NMR showed
the IBUsorbitol in-house standard is racemic mixture of four
isomers, mainly sorbit-1-yl 2-(4-isobutylphenyl)propionate and The characterization of reference standards
sorbit-6-yl 2-(4-isobutylphenyl)propionate. Each NMR 1H and 13C The obtained MS data for prepared standards were in agreement
NMR signal of IBUsorbitan was formed by one major, one with the data measured in IBU SGCs pharmaceutical formulation
medium and 12 minor peak due to mixture of four isomers, mainly (Fig. 5). UHPLC peak equivalency between synthesized in-house
1,4-anhydrosorbit-6-yl 2-(4-isobutylphenyl)propionate and 3,6- standards and impurities eluting in stability and forced degradation
anhydrosorbit-1-yl 2-(4-isobutylphenyl)propionate. Further stereo- samples was also conrmed (data not shown).
chemical investigation of the formed monoesters was not subject of Finally, esters were quantitatively characterized for pharmaceuti-
this study. cal use and the corresponding in-house reference standards were es-
tablished. The potency of the standards was calculated based on the
values obtained from determination of impurities (organic, inor-
Discussion
ganic, water and residual solvents) by applying the principle of mass
Identication of unknown degradation products balance. The potency was 97.0% for IBUPEG monoester, 95.3%
The routine UPLC method was then optimized to reach separation for IBUsorbitol and 90.9% for IBUsorbitan. The response correc-
of IBUPEG oligomers. Previously, the resolution between PEG oli- tion factor was also determined in respect to IBU API for UHPLC
gomers using LC methods was achieved at elevated temperature UV method at 220 nm, namely 3.7 for IBUPEG monoester, 1.9 for
using gradient elution (30). The routine UHPLC method was IBUsorbitol and 1.9 for IBUsorbitan.
Figure 5. Full scan HRAM mass spectra in positive APCI ionization mode of synthesized IBUPEG monoester (A), IBUPEG diester (B, average mass spectrum at
RT ranging from 24.0 to 25.9 min), IBUsorbitol (C) and IBUsorbitan (D).
8 Doua et al.