DLP drug used for investigation was obtained from the Research & Development Department of Dr. Reddy’s Laboratories Ltd.,Hyderabad,India. HPLC grade acetonitrile,triethyl amine,ortho phosphoric acid,hydro chloric acid,sodium hydroxide,hydrogen peroxide,ammonium acetate,ammonia and potassium dihydrogen phosphate were obtained from Merck Co.,Mumbai,India. Ultra-pure water was collected from TKA Millipore water purification system. Dimethyl sulfoxide (DMSO)-d₆ (for NMR) was obtained from Aldrich Chemical Co.,USA.

The chromatography analysis was performed using a Waters Alliance 2695 separation module (Waters Corporation,Milford,USA) equipped with a 2998 Photo Diode Array (PDA) Detector,a degasser,a quaternary pump,and an auto sampler system. The output signal was monitored and processed using Empower 2 software. Preparative isolation work performed on an Agilent 1200 series preparative HPLC system which was equipped with an automated fraction collector and a photodiode array detector. The data was collected and processed using Chemstation software. Rotavapour model: Heidolph Laboratory 4002 was used for solvent evaporation. Lyophilizer (model: Virtis Advantage plus) was used to solidify the impurities. The electro spray ionization and MS/MS studies were performed on triple quadrupole mass spectrometer AB SCIEX model: 4000 QTRAP. The UPLC-MS/TOF system consisted of an ACQUITY^TM Ultra Performance Liquid Chromatography (UPLC) system operating in either positive or negative ion mode with electrospray ionisation (ESI) source. NMR experiments (1D and 2D) were performed using a 500 MHz Unity INOVA NMR spectrometer (Varian). Cintex digital water bath was used for hydrolysis studies. Photo stability studies were carried out in photo stability chamber (Sanyo,Leicestershire,UK). Thermal stability studies were performed in a dry air oven (Cintex,Mumbai,India). A pH meter (Thermo Scientific) was used to check and adjust the pH of buffer solutions.

Forced degradation studies were performed on the DLP drug with the intention to study the stability of DLP under different conditions and to explore the degradation products. Various trials were taken to optimize the degradation conditions. For acid,base hydrolysis and oxidative study 100 mg of DLP was dissolved in 10 mL of diluent (0.1 mol/L NaOH-acetonitrile,75:25 (v/v)),then 10 mL of stressor (0.1 mol/L HCl,0.1 mol/L NaOH,10% (v/v) H₂O₂) was added to it and refluxed at 40-60 ℃,then neutralized and made up to 100 mL with diluent. For thermal stress study,the drug was kept at 105 ℃ for 24 hours. Photolytic studies were carried out by exposing DLP and its solutions in a photo stability chamber at 1.2 million lux hours and UV light at 200 Watt hours/meter². Parallel blanks were kept under all the stress conditions. The optimized stress conditions employed are enlisted in Table 2.

Table 2

Table 2(Table 2)

Table 2 Optimized stress conditions of DLP degradation study Degradation study Degradation conditions Time Acid hydrolysis 0.1 mol/LHCl at 60 t 5 min Base hydrolysis 0.1 mol/LNaOHat 60 t 10 min Water hydrolysis water at 60 t 30 min Oxidative hydrolysis 10% (v/v) H2O2 on bench top 30 min Thermal 105 t 24 h Photolytic 1.2 million lux hours of visible light and 200 Watt hours/meter2 of UV light 48 h

Table 2 Optimized stress conditions of DLP degradation study

HPLC trails were taken to develop a stability indicating method for the separation and quantification of DLP and DPs using different buffers,pH conditions and different proportions of acetonitrile and methanol. Xterra RP18 column (250 mm×4.6 mm,3.5 μm) was used for separation. The optimized HPLC conditions were described in Table 3. The stressed samples were analyzed by HPLC method to monitor the formation of different DPs under various degradation conditions. DLP spiked with all impurities are shown in Fig. 1.

Table 3

Table 3(Table 3)

Table 3 LC conditions for analytical, preparative, LC-MS/MS and UPLC-MS/TOF analyses Technique Mobile phases Flow rate/(mL/min) Gradient program time/min (B/%) Injection volume/μL A B HPLC buffer* -ACN A-ACN 0.8 0(10)； 50 (60)；67-85 (80) 5.0 (90:10) (30:70) 90-100 (10) Preparative 0.1%NH3·h2o A-ACN 6 0 (40) ； 40 (40)；41-50 (100) 500.0 (30:70) 51 (0) ； 60 (0) LC-MS/MS buffer** -ACN buffer* -ACN 1 0(10)； 50 (60)；67-85 (80) 5.0 (90:10) (30:70) 90-100 (10) UPLC-MS/TOF buffer** -ACN ACN-methanol 0.24 0-2 (6)； 4-5 (12)； 16 (40) 1.0 (95:5) (60:40) 25-33 (55)； 38-42 (90)； 43-50 (6) * 10 mL of triethyl amine was added to 100 mL of 0. 01 mol/L KH2 PO4 buffer and pH was adjusted to 9. 0 using orthophosphoric acid. ** 0. 01 mol/L NH4OOCCH3 buffer with the pH adjusted to 9. 0 using 1% ammonia solution. Column oven temperature, sample temperature and detection wavelength were 30 ℃ , 5 ℃ and 285 nm, respectively. All the fractions were volume ratios.

Table 3 LC conditions for analytical, preparative, LC-MS/MS and UPLC-MS/TOF analyses

Fig. 1

Fig. 1 Chromatogram of DLP spiked with all impurities

The stressed samples were targeted to preparative HPLC analysis with the objective to isolate oxidative DP (RRT 0.54 (DP Ⅰ)). The preparative method was developed by employing a mobile phase at basic pH on Inertsil ODS-3 column (250 mm×10 mm,5.0 μm). The optimized preparative LC conditions are described in Table 3. The impurity fractions were collected separately from several injections and pooled together after confirming the purities of each fraction. The solvent present in the pooled fraction was evaporated by using rotavapour under high vacuum. The aqueous solutions were lyophilized to solidify the impurities.

MS/TOF and MS/MS studies were performed in ESI positive mode. The instrument parameters were first optimized to get the molecular ion peak of the DP. The same were subsequently modified to get complete fragmentation profile of purified DP. In a further step,the information on the origin of each individual fragment was obtained from MS² studies.

The ESI mass spectrum of the impurity was recorded on AB SCIEX 4000 QTRAP LC-MS/MS system. The samples were introduced into the system at the ion spray voltage of 5 500 V and the temperature was maintained at 400 ℃. GS1 and GS2 were optimized to 30 psi (206.85 kPa) and 35 psi (241.325 kPa) respectively and the declustering potential was 70 V. A mixture of water and acetonitrile in the proportion of 20:80 (v/v) was used as diluent for sample preparation and mass concentration of sample was 0.02 mg/mL. Analysis was carried out on Xterra RP18 column (250 mm×4.6 mm,3.5 μm) column. Based on the optimized HPLC conditions an MS compatible HPLC method was developed. The optimized LC conditions are described in Table 3.

The analysis was acquired using the lock spray to ensure accuracy and reproducibility. High resolution (wavelength (W) mode,full width half maximum (FWHM) 10 500) positive polarity scan responses were collected from m/z 100 to 1 000 at a rate of 1.0 s/scan. The source block and desolvation temperatures were 90 ℃ and 180 ℃,respectively. The nebulizer and desolvation gas (nitrogen) flows were 20 L/h and 450 L/h,respectively. The instrument parameters were used as capillary 3 000 V,cone 25 V,extractor 2 V and microchannel plate (MCP) 2 700 V. The acquisitions were done in scan mode. MS/TOF studies were performed in ESI positive mode. A mixture of water and acetonitrile in the proportion of 20:80 (v/v) was used as diluent for sample preparation and the mass concentration of sample was 0.02 mg/mL. ACQUITY UPLC^TM BEH shield,RP18 column (100 mm×2.1 mm,1.7 μm) was used for separation. Based on MS compatible HPLC method,a UPLC method was developed. The optimized LC conditions are described in Table 3.

NMR experiments (1D and 2D) were performed in DMSO-d₆ at 25 ℃ as solvent. ¹H NMR measurements were carried out at 500 MHz,while ¹³C NMR experiments were performed at 125 MHz. Proton and carbon chemical shifts were reported on δ scale in ppm,relative to tetra methyl silane (TMS) (δ=0.00 ppm (10^-6)) and DMSO (δ=39.50 ppm) as internal standards in ¹H and ¹³C NMR spectra respectively. Standard pulse sequences provided by Varian were used for 1D and 2D NMR data.

Forced degradation is degradation of a drug substance or a drug product under conditions more severe than the accelerated stability conditions. Degradation studies were performed on DLP drug substance with an objective to establish the DPs of the drug substances,structure elucidation and intrinsic stability of DLP. The stress factors used for forced degradation studies included acid,base,neutral hydrolysis,thermal degradation,photolysis and oxidation. Hence there is no specification in regulatory guidelines about the stress conditions of pH,temperature and specific oxidizing agents to be used. We put efforts to optimize the degradation conditions to evaluate the degradation behavior of DLP. In all stress studies,trails were taken from mild to extreme stress conditions i. e. higher temperature,higher concentrations of acid,base and peroxide and prolonged exposure times to optimize the degradation conditions to obtain reasonable degradation of DLP. The optimized stress conditions are listed in Table 2. All these stressed samples were subjected to HPLC analysis.

All the stressed samples were analyzed by aforementioned HPLC method. In newly developed stability indicating HPLC method all the DPs and DLP were well separated. HPLC analysis revealed that the drug was highly unstable under acid,base,oxidative and thermal conditions and formed known and unknown impurities. It was moderately stable under photolytic degradation (Fig. 2). This complete degradation study also inferred that a small change in the stress conditions like concentration of stressors and exposure times resulted in a different degradation profile including type of formation of DPs and their percentages. The comparisons of formation of known and unknown DPs under different stress conditions by HPLC study are shown in Table 4.

Fig. 2

Fig. 2 Overlay chromatograms of forced degradation studies of DLP

Table 4

Table 4(Table 4)

Table 4 Comparisons of formation of degradants in stress study Item Percentages of degradants/% Imp A Imp B Imp C Imp D Imp E Imp F Imp G Imp Ⅰ Imp Ⅱ Imp Ⅲ Imp M+467 DP Ⅰ RRT 0.14 0.25 0.36 0.71 0.74 1.06 1.43 1.41 1.65 1.68 1.85 0.54 Control sample 0.01 0.00 0.01 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Acid degradation 1.51 2.01 3.56 0.00 0.00 0.00 0.0 10.05 0.00 0.00 000 3.02 Base degradation 0.22 0.02 1.65 0.00 0.00 0.02 0.03 0.1 0.35 0.01 0.00 0.00 Peroxide degradation 3.10 8.50 0.00 0.00 0.00 0.11 0.00 5.23 0.00 0.00 0.00 6.35 Thermal degradation 0.62 2.82 8.63 0.00 0.58 2.32 0.59 3.20 0.51 0.62 6.47 1.56 Hydrolytic degradation 0.24 0.15 0.01 0.00 0.00 0.12 0.21 1.23 0.25 0.86 0.00 0.00 Photolytic degradation 0.00 0.01 0.21 0.00 0.00 0.02 0.00 0.02 0.00 0.00 0.00 0.00

Table 4 Comparisons of formation of degradants in stress study

Isolation of impurities (DPⅠ) by preparative HPLC was a challenging job as the target impurities were highly labile. It is indeed a major task to stabilize the impurities during its isolation. Hence it was observed that DPⅠ is very unstable to acidic pH,simple and efficient preparative HPLC method was developed by employing a basic pH mobile phase using ammonia as pH modifier. Several trials were performed to optimize the purity,yield and throughput. Finally a simple gradient system was developed on an Inertsil ODS-3 column (250 mm×10 mm,5.0 μm) with good resolution and a run time of 60 min. The chromatographic purity of DPⅠ was 99.1 and this sample was used for spectroscopic structural studies. The structures of DPⅠ and DLP are shown in Fig. 3.

Fig. 3

Fig. 3 Chemical structures of DLP and DPⅠ with numbering

The ESI-MS spectra of DPⅠ show only positive ionization and the molecular ion is observed at m/z 322.1 (Fig 4). The compound didn’t form lithium adduct ion and negative ESI-MS spectrum shows no peaks,suggesting that the molecular ion obtained is positively charged [25],i. e.,m/z 322.1 due to M⁺. The positive HR-MS spectrum shows peak at m/z 322.115 4 (Fig. 4),corresponding to a possible molecular formula C₁₆H₁₅N₃F₃O. In comparison,DPⅠ contains one hydrogen more and one oxygen and one sulphur atoms less than DLP. This indicates the elimination of SO (sulphur monoxide) moiety from the parent DLP and suggests rearrangement of benzimidazole and pyridine moieties to form DPⅠ. The plausible degradation pathways for DPⅠ are shown in Fig. 5. The comparisons of MS/MS spectra of DLP and DPⅠ don’t show any common fragment ions.

Fig. 4

Fig. 4 HR MS and MS/MS data of DLP and DPⅠ a. MS spectrum of DLP; b. MS/MS spectrum of DLP; c. MS spectrum of DPⅠ; d. MS/MS spectrum of DPⅠ.

Fig. 5

Fig. 5 Plausible degradation pathways for DPⅠ

The comparisons of 1H NMR spectra of DPⅠ with DLP (Fig. 6) show the absence of one methylene moiety (in DLP 4.83 & 4.76 ppm,Ha,Hb) and addition of one methyl signal (at 2.66 ppm,Table 5). The absent methylene moiety was found to be the one attached to sulphur in DLP. The benzimidazole,pyridine and trifluoro ethoxy methylene protons were found to be intact in DPⅠ. The methyl signal at 2.24 ppm (corresponding to C18) showed nuclear overhauser effect spectroscopy (NOESY) correlations with the methylene protons at 5.31 ppm and methyl protons at 2.66 ppm. The NOESY (Fig. 7) and heteronuclear multiple bond correlation (HMBC) (Fig. 8) correlations confirm that the methyl is substituted at ortho position to the methyl at 2.24 ppm on pyridine moiety.

Fig. 6

Fig. 6 NMR spectra of DLP and DPⅠ

Table 5

Table 5(Table 5)

Table 5 1H and 13C NMR assignments for DLP and DP Ⅰ Position* Dexlansoprazole DPⅠ 1H δ/ppm J**/Hz 13C DEPT 1H δ/ppm J/Hz 13C DEPT 1 NH 13.57/s - - - NH - - - - 2 - - - 154.15 - - - - 155.8 - 3 N - - - - N - - - - 4 2H (4, 7) 7.72, 7.59/br - 119.72, 112.35 CH 2H (4, 7) 7.45/m - 117 CH 5 2H (5,6) 7.32, 7.30/br - 123.85, 122.50 CH 2H (5, 6) 6.94/m - 118.56 CH 6# 119.72, 112.35 CH 117 CH 7# 123.85, 112.50 CH 118.56 CH 8# - - - 134.63 - - - - 145.73 9# 143.03 - 145.73 10 S - - - - - - - - - 11 2H, (Ha,Hb) 4.76, 4.83 14.5 59.97 CH2 3H 2.66/s - 18.27 CH3 12 N - - - - N - - - - 13 - - - 150.91 - - - - 154.38 - 14 - - - 121.94 - - - - 124.5 - 15 - - - 161.27 - - - - 165.6 - 16 1H 7.10/d 5.0 107.01 CH 1H 7.70/d 7.0 108 CH 17 1H 8.29/d 5.0 148.09 CH 1H 9.12/d 7.0 145.18 CH 18 3H 2.19/s - 10.52 CH3 3H 2.24/s - 11.27 CH3 19 O - - - - O - - - - 20 2H 4.90/q 8.5 64.32/q(34.4 Hz) CH2 2H 5.31/q 8.5 65.95/q (35.0 Hz) CH2 21 - - - 123.74/q(276 Hz) - - - - 123.3/q (280 Hz) - * The numberings are same as in Fig. 1. ** 1H-1H coupling constant; # interchangeable. s, singlet; d, doublet; t, triplet; q, quar-tet; m, multiplet, br, broad.

Table 5 1H and 13C NMR assignments for DLP and DP Ⅰ

Fig. 7

Fig. 7 NOESY spectrum of DPⅠ

Fig. 8

Fig. 8 HMBC spectrum of DPⅠ

This confirms that the methylene moiety in DLP has changed to a methyl moiety in DPⅠ. Based on double bond equivalence and the positively charged properties,the nitrogen in pyridine moiety in DPⅠ directly connected to the benzimidazole moiety with positive charge.Based on the above spectral data,the structure of DPⅠ was characterized as 1-(1H-benzo[d]imidazol-2-yl)-2,3-dimethyl-4-(2,2,2-trifluoroethoxy) pyridin-1-ium.

The drug substance of dexlansoprazole was subjected to stress studies and evaluated for degradation products. The unknown oxidation degradation products were isolated by a simple preparative method and characterized by NMR,MS/MS and MS/TOF spectroscopic techniques. The most probable structure was proposed for impurities based on the spectral data. The entire study will benefit in developing stable drug substances and drug products,and in recommending storage conditions and shelf lives.

Acknowledgement We would like to thank the management of Dr. Reddy’s Laboratories Ltd. for supporting this work. Cooperation from colleagues of Research & Development Department and Analytical Research & Development Department of Dr. Reddy’s Laboratories Ltd. is acknowledged.