INTRODUCTION:

The present investigation was undertaken to establish and validate a robust, selective and highly sensitive LC-MS/MS analytical method for the determination of N-nitrosodibutylamine (NDBA), a potential nitrosamine impurity, in Simvastatin and Ezetimibe tablet formulations to ensure drug safety, quality, and regulatory compliance⁴. Simvastatin and Ezetimibe are widely prescribed lipid-lowering medications used either individually or in fixed-dose combination therapy for the management of hypercholesterolemia and prevention of cardiovascular diseases⁵. Simvastatin belongs to the statin class of drugs and acts by inhibiting HMG-CoA reductase, the rate-limiting enzyme responsible for cholesterol biosynthesis in the liver, thereby reducing plasma cholesterol levels⁶. Ezetimibe works through a different mechanism by selectively inhibiting the intestinal absorption of cholesterol from the small intestine⁷. The complementary mechanisms of these two agents make their combination particularly effective in reducing low-density lipoprotein cholesterol (LDL-C) and lowering cardiovascular risk^5,17. In recent years, nitrosamine impurities in drug products have become a significant safety concern for both regulatory agencies and the pharmaceutical industry. Nitrosamines are a class of highly potent genotoxic impurities widely recognized for their mutagenic and carcinogenic properties²¹. Several compounds belonging to this class have been classified as probable human carcinogens. Among them, NDBA is an N-nitrosamine compound that may occur as an impurity in pharmaceutical drug substances or finished drug products. The presence of NDBA and other nitrosamine impurities in pharmaceutical products poses a potential risk to patient safety and may compromise the overall quality of drug products¹⁷. Even trace amounts of these impurities can increase the long-term carcinogenic risk associated with continuous drug exposure⁸. As a result, regulatory agencies have established strict acceptable intake limits for nitrosamines in pharmaceutical formulations to ensure patient safety and product quality. Nitrosamines such as NDBA are typically formed through reactions between secondary or tertiary amines and nitrosating agents (e.g., nitrites or nitrogen oxides) under specific favorable conditions, including acidic pH (commonly pH 2–5), elevated temperatures, the presence of oxidizing or catalytic species, and environments with sufficient nitrite availability and low oxygen scavenging capacity, all of which promote the generation of reactive nitrosating intermediates (such as N₂O₃) that facilitate nitrosation¹⁸. These reactions may occur during different stages of pharmaceutical manufacturing including drug substance synthesis, formulation development, packaging, or long-term storage²². Factors such as the presence of amine-containing intermediates or reagents, nitrite impurities in excipients, acidic reaction conditions, elevated temperatures, and prolonged storage may promote the formation of nitrosamines⁹. In the case of Simvastatin and Ezetimibe, the use of specific reagents or intermediates such as secondary amines (e.g., diethylamine, dibutylamine), tertiary amines (e.g., triethylamine), and nitrite-containing reagents (e.g., sodium nitrite) or nitrosating species formed in situ under acidic conditions during the manufacturing process may create a potential pathway for the formation of nitrosamine impurities such as NDBA. Due to the potential carcinogenic risk associated with nitrosamines, regulatory authorities including the United States Food and Drug Administration (USFDA), European Medicines Agency (EMA), and World Health Organization (WHO) have reported the detection of nitrosamine impurities in several pharmaceutical products^1,4. These findings have led to a global reassessment of pharmaceutical manufacturing processes, resulting in product recalls and the implementation of strict regulatory guidance for the control of nitrosamine impurities^21,22. Accordingly, pharmaceutical industries must carry out thorough risk assessments and establish reliable analytical methods for monitoring and quantifying nitrosamines in both active pharmaceutical ingredients and finished dosage forms¹⁹. The analysis of nitrosamine drug substance-related impurities (NDSRIs) specifically requires highly sensitive and selective techniques capable of detecting trace-level contaminants within complex pharmaceutical matrices¹. Various analytical approaches have been utilized for nitrosamine analysis, such as GC–MS, LC–MS, and LC–MS/MS. Of these, LC–MS/MS is widely preferred owing to its high sensitivity, selectivity, and specificity, which allow reliable quantification of nitrosamines at trace (ppb) levels^10,14. Hence, it is crucial to develop a robust and validated analytical procedure to ensure accurate detection and quantification of nitrosamine impurities in pharmaceutical formulations. Method validation establishes the reliability and reproducibility of the procedure by evaluating key parameters including specificity, linearity, accuracy, precision, limit of detection, and limit of quantification, as outlined in the International Council for Harmonisation (ICH) guidelines¹¹. The validated LC–MS/MS method presented in this work offers a reliable and efficient approach for routine quality control and regulatory surveillance of NDBA in Simvastatin and Ezetimibe tablet formulations¹⁵.

Figure 1. Acid-catalyzed N-nitrosation of secondary amines by nitrite leading to N-nitrosamine formation.

2. MATERIALS AND METHODS:

2.1 Chemicals and reagents:

Ezetimibe and Simvastatin tablet formulations and associated impurities were provided by Ikon Remedies Pvt Ltd, Mumbai, Maharastra, India, India used for analysis. All aqueous solutions, including the mobile phase, were prepared using HPLC grade water obtained from a Milli-Q purification system (Merck Millipore). Methanol (LC-MS grade) was obtained from Merck (Germany), while formic acid (HPLC grade) was purchased from Quiligens. All chemicals and solvents used in the present study were of analytical, HPLC, or LC–MS grade and were used without further purification.

2.2 Instrumentation:

Chromatographic separation was performed using an ACQUITY Premier UPLC system (Waters Corporation, Milford, MA, USA) equipped with a photodiode array detector (PDA) operated at 230nm and coupled to a Xevo TQ-S Micro Triple Quadrupole Mass Spectrometer with an Atmospheric Pressure Chemical Ionization (APCI) source. The system was controlled, and data acquisition and processing were carried out using MassLynx software (Waters Corporation, Milford, MA, USA)^16,20.

2.3 Chromatographic Conditions:

Separation was achieved using a Zorbax SB-Phenyl column (150 × 4.6mm, 3.5µm; Agilent Technologies). The mobile phase comprised 0.1% (v/v) formic acid in water (A) and 0.1% (v/v) formic acid in methanol (B), with gradient elution applied to resolve the analytes. The system was operated at a flow rate of 0.5mL/min, and the column temperature was maintained at 40°C. A 10 µL injection volume was used, with a total analysis time of 20minutes. Detection was performed using a PDA detector at 230nm. Several gradient conditions were evaluated to optimize separation performance².

2.4 Mass Spectrometric Conditions:

Detection was carried out using a Xevo TQ-S Micro triple quadrupole mass spectrometer (Waters Corporation, Milford, MA, USA) fitted with an APCI source operating in positive ionization mode. Data acquisition was performed in multiple reaction monitoring (MRM) mode to achieve high sensitivity and selectivity for N-nitrosodibutylamine (NDBA)²⁰. The source conditions were optimized, with a corona voltage of 3.0kV and a cone voltage of 40V. The source temperature was set at 120°C, while the APCI probe was maintained at 250°C. The desolvation gas flow was maintained at 700L/h, and the cone gas flow was set at 20L/h. The quadrupole resolution parameters were optimized with LM resolution set at 8.1 and HM resolution at 15.0. The ion energy parameters were adjusted with ion energy-1 at –0.7 and ion energy-2 at 0.3 to achieve optimal ion transmission and sensitivity. The analysis was performed in MRM mode with a mass span of 0.5 Da, and the retention window was set from 4.0 to 12.0 minutes for monitoring the target analyte.

2.5 Preparation of Standard Solutions:

NDBA Impurity Stock Solution:

An NDBA impurity stock solution was prepared at a concentration of 1000ppm by accurately weighing an appropriate amount of NDBA and dissolving it in methanol. The solution was thoroughly mixed to achieve complete dissolution of the components.

NDBA Working Stock Solution:

A working stock solution of 1ppm was prepared by appropriate dilution of the NDBA impurity stock solution using the diluent.

2.6 Sample Preparation:

The tablets were crushed to obtain a uniform powder. An amount equivalent to 10mg of Ezetimibe was transferred into a 25mL volumetric flask, and about 16 mL of diluent (water: methanol, 1:1 v/v) was added. The mixture was sonicated for 10minutes, diluted to volume with diluent, and mixed well. The solution was then filtered through a 0.22µm PVDF syringe filter to obtain a 400ppm solution.

3. METHOD DEVELOPMENT AND OPTIMIZATION:

3.1 Optimization of Chromatographic Conditions:

The chromatographic parameters were carefully optimized to ensure effective resolution of the NDBA impurity from the active pharmaceutical ingredients, while also achieving satisfactory peak symmetry, analytical sensitivity, and method reproducibility¹⁵. Several trials were performed by modifying column chemistry and gradient programs. For all chromatographic experiments, the mobile phase comprised 0.1% formic acid in water as solvent A and 0.1% formic acid in methanol as solvent B. The analyses were performed at a constant flow rate of 0.5mL/min with the column oven maintained at 40°C. A sample injection volume of 10 µL was used, and chromatograms were monitored at a wavelength of 230 nm. Multiple chromatographic conditions were systematically investigated to assess the influence of gradient elution profiles and column type on separation efficiency and overall chromatographic performance¹³. The details of the chromatographic trials and their respective objectives are summarized in Table 1.

3.2 Optimization of Mass Spectrometric Parameters:

The mass spectrometric settings were carefully optimized to enable highly sensitive and selective detection of the NDBA impurity. The study employed atmospheric pressure chemical ionization (APCI) operated in the positive ionization mode for the analysis, and tuning experiments were carried out by adjusting key source parameters to obtain stable ionization and adequate signal intensity¹². The finalized operating parameters consisted of a corona current set at 2.5 kV, a cone voltage of 40V, and a source temperature maintained at 120°C. Quantitative analysis was carried out using MRM mode to ensure selective detection of the analyte as shown in table 2. Mass spectrometric data were acquired in the positive ionization mode, using a mass span of 0.5 Da, and data acquisition was restricted to a retention time window of 4.0–12.0min to enhance sensitivity and reduce background interference during LC-MS/MS analysis. Flow diversion program: 4.0min – LC; 6.0min – Waste; 9.0min – LC; 12.0min – Waste. The gradient program used for LC–MS/MS analysis was adopted from the optimized chromatographic conditions obtained in Trial 4 of the LC method development. Ionization was carried out using APCI in positive mode, and multiple reaction monitoring (MRM) transitions were investigated based on the precursor ion at m/z 159. The parent ion at m/z 159 produced several product ions during fragmentation, among which the product ion at m/z 103 showed optimal response and peak shape; therefore, the MRM transition 159 → 103 was selected for quantification of the NDBA impurity¹². A stock solution of NDBA (1000ppm) was prepared in methanol, from which a 5-ppm working solution was subsequently obtained using the selected diluent. The sample solution was prepared at a concentration of 10,000ppm in the same diluent. To identify an appropriate diluent, different options were evaluated, including methanol (Trial 1), water (Trial 2), and a 1:1 mixture of water and methanol (Trial 3). Different diluent systems were also evaluated during method development. Methanol showed interference in the chromatogram, while water resulted in suppression of the NDBA response. A binary solvent system composed of water and methanol (50:50, v/v) provided satisfactory response and recovery without interference and was therefore selected as the optimized diluent for sample preparation.

Table 2. Evaluated MRM Transitions for NDBA Impurity

Compound	Parent (m/z)	Product Ion (m/z)	Dwell Time (s)	Cone Voltage (V)	Collision Energy (V)
NDBA	159.0	57	0.1	40	12
NDBA	159.0	103	0.1	40	10

Figure 2. Proposed fragmentation pathway of NDBA showing the formation of product ions from the parent ion (m/z 159) used for MRM transitions.

3.3 Selection of Final optimized LC-MS/MS conditions:

Separation of the analytes was carried out on a Zorbax SB-Phenyl column with a modified gradient program, providing adequate resolution between NDBA and ezetimibe, which eluted at approximately 11.7min and 12.3min, respectively. These conditions were selected as the optimized LC parameters for further analysis. Mass spectrometric analysis was conducted using APCI operating in the positive ion mode. The precursor ion at m/z 159 was selected, and the MRM transition 159 → 103 was used for quantification based on its optimal response and peak shape. A 50:50 (%v/v) mixture of water and methanol was chosen as the diluent for the study, as it provided satisfactory analyte response and recovery without chromatographic interference.

Figure 3. Mass fragmentation pattern of NDBA

4. METHOD VALIDATION:

The analytical method validation was performed in accordance with the principles outlined in the ICH Q2(R2) Analytical Method Validation guideline for validation of analytical procedures³. The validation strategy was designed to demonstrate that the proposed LC-MS/MS method is suitable for its intended purpose by evaluating critical performance characteristics, including specificity, linearity, accuracy, precision, limit of detection (LOD), limit of quantification (LOQ), and robustness^10,3. All experiments were conducted to ensure compliance with internationally accepted regulatory requirements for analytical method validation in pharmaceutical analysis.

4.1 Specificity:

Method specificity was assessed by injecting blank, diluent, filtered diluent and sample preparations to verify the absence of any interfering peaks at the retention time corresponding to NDBA³.

4.2 Linearity:

Linearity for NDBA was assessed across a concentration range extending from the LOQ level up to 150% of the target concentration. Calibration standards were prepared at LOQ, 25%, 50%, 75%, 100%, 125%, and 150% levels. A calibration plot was generated by correlating analyte concentration with the corresponding peak area to assess the linear response of the detector across the studied range.

4.3 Limit of Detection (LOD) and Limit of Quantification (LOQ):

The LOD and LOQ were established using replicate injections at low concentration levels with evaluation of the signal-to-noise ratio and peak response. The LOD was established to assess the method sensitivity for detecting trace levels of NDBA, while the LOQ was determined to confirm reliable quantification at the lowest concentration with acceptable precision and detector response³.

4.4 Accuracy (Recovery Study):

Method accuracy was assessed by performing recovery experiments at the LOQ level and at 50%, 100% and 150% concentration. Three sets were analyzed at each level, and the mean recovery of the three sets was calculated to assess the accuracy of the method.

4.5 Precision:

Method precision was determined by analysing six independently prepared sample replicates under identical experimental conditions. Intermediate precision was evaluated using independent preparations under different conditions, while intermediate system precision was checked through replicate standard injections to confirm instrument consistency. Overall precision was evaluated by integrating the results from method precision and intermediate precision studies to determine the method’s combined variability and performance.

4.6 Robustness:

The robustness of the developed method was assessed by deliberately varying selected chromatographic and instrumental conditions, such as flow rate, probe temperature, and desolvation gas flow, to evaluate their impact on method performance. The effect of these variations was assessed in terms of system suitability, % recovery, and percentage difference values. The study was done to evaluate the method’s reliability and consistency when subjected to minor variations in analytical conditions³.

5. RESULTS AND DISCUSSION:

5.1 Specificity:

No interfering peaks were detected from the diluent or filtered diluent at the retention time of NDBA, demonstrating the specificity of the method for its determination.

5.2 Linearity:

The method exhibited excellent linearity over the investigated concentration range, with a correlation coefficient (R) of 0.9991 and a coefficient of determination (R²) of 0.9981. The calibration curve was described by the regression equation y = 4,641,715.3867x − 5.5993, where the slope was 4,641,715.3867 and the y-intercept was −5.5993. The low intercept value indicates negligible systematic bias and confirms a strong linear relationship between NDBA concentration and detector response throughout the studied range, as shown in Figure 4.

Figure 4. Linearity Plot of Concentration (µg/mL) v/s Peak area of NDBA

5.3 Limit of Detection (LOD) and Limit of Quantification (LOQ):

The sensitivity of the developed LC–MS/MS method was evaluated by establishing the limit of detection (LOD) and limit of quantification (LOQ) through six replicate injections at each concentration level. Following the signal-to-noise (S/N) approach described in ICH Q2(R2), the lowest detectable and quantifiable concentrations of NDBA were assessed. At the LOD level, the average peak response was 144 with a mean S/N ratio of 9.49, while the corresponding %RSD was 3.8%, demonstrating consistent detection of the analyte at trace levels. For the LOQ study, an average peak area of 484 and a mean S/N ratio of 29.04 were obtained, with a %RSD of 6.3%, indicating satisfactory repeatability for quantitative measurement³. These results confirm that the method is sufficiently sensitive for the reliable detection and quantification of NDBA in the investigated pharmaceutical formulation.

Table 3. Summary of LOD and LOQ results for NDBA impurity

Parameter	LOD	LOQ
Mean Peak Area	144	484
Mean S/N Ratio	9.49	29.04
%RSD (Peak Area)	3.8	6.3

5.4 Accuracy (Recovery Study):

Method accuracy was determined through recovery experiments conducted at the LOQ level as well as at 50%, 100%, and 150% concentration levels. The predefined acceptance criterion was a mean recovery within 80.0-120.0% at each level, with %RSD not exceeding 20.0%. The mean recoveries obtained were 112.8%, 105.9%, 100.5%, and 107.1% at the LOQ, 50%, 100%, and 150% levels, respectively. The corresponding %RSD values were 12.7%, 1.3%, 5.5%, and 1.8%. All results met the acceptance criteria, confirming the accuracy of the method across the studied concentration range.

Table 4. Summary of Accuracy results for NDBA impurity

Level	LOQ	50%	100%	150%
Mean	112.8	105.9	100.5	107.1
%RSD	12.7	1.3	5.5	1.8

5.5 Precision:

Precision of the developed LC–MS/MS method was investigated by evaluating system precision, method precision, intermediate precision, and combined precision. System precision was assessed using six replicate injections of the standard solution (n=6), whereas method precision was determined from six independently prepared sample solutions (n=6). Intermediate precision was established by analyzing six separate sample preparations under varying analytical conditions (n=6). The overall precision was further assessed by combining the results obtained from method precision and intermediate precision studies (n=12). The %RSD values observed for system precision, method precision, intermediate precision, and combined precision were 1.4%, 3.9%, 2.8%, and 3.3%, respectively. All values were below the predefined acceptance criterion of not more than 20.0%, indicating good agreement among replicate measurements. These findings demonstrate that the proposed method provides consistent and reproducible results and is suitable for the routine determination of NDBA impurity in Simvastatin and Ezetimibe tablet formulations.

Table 5. Summary of precision results for NDBA impurity

Parameter

System Precision

Method Precision

Inter-mediate

Precision

Method Intermediate system Precision

Mean

4715

101.4

100.1

100.6

%RSD (NMT 20%)

1.4

3.9

2.8

3.3

5.6 Robustness:

The robustness evaluation indicated that minor intentional variations in flow rate, probe temperature, and desolvation gas flow had no significant impact on the overall performance of the method. The %RSD values were found within 1.3%–3.1%, while % recovery ranged from 97.6% to 105.5% under all varied conditions. The obtained percentage difference values were within acceptable limits, confirming that the developed method is robust and reliable for routine analysis.

Table 6. Robustness study showing the effect of variations in flow rate, probe temperature, and desolvation gas flow on %RSD values of the developed analytical method

Flow rate	0.4 mL/minute	0.5 mL/minute	0.6 mL/minute
% RSD	2.3	1.4	1.7
Probe temperature	200°C	250°C	300°C
% RSD	1.3	1.4	3.1
Desolvation gas flow	650 L/Hr	700 L/Hr	750 L/Hr
% RSD	1.9	1.4	2.5

Table 7. Robustness study showing %recovery and %difference under varied analytical conditions

Flow rate	0.4 mL/minute	0.5 mL/minute	0.6 mL/minute
% Recovery	103.2	101.4	99.1
% Difference (NMT)	1.78	NA	2.27
Probe temperature	200°C	250°C	300°C
% Recovery	98.7	101.4	105.5
% Difference (NMT)	2.66	NA	4.04
Desolvation gas flow	650 L/Hr	700 L/Hr	750 L/Hr
% Recovery	102.6	101.4	97.6
% Difference (NMT)	1.18	NA	3.75

6. CONCLUSION:

A robust, sensitive, and selective LC–MS/MS method was developed and validated for the quantification of N-nitrosodibutylamine (NDBA) impurity in Simvastatin and Ezetimibe tablet dosage forms. The optimized chromatographic and mass spectrometric parameters enabled efficient separation and specific detection of the target impurity with satisfactory signal response. Validation was carried out in accordance with ICH guidelines, and the method exhibited acceptable performance in terms of specificity, linearity, accuracy, precision, sensitivity, and reproducibility. Excellent linear behavior was observed across the evaluated concentration range, along with low limits of detection and quantification, highlighting the method’s capability for trace-level analysis. Recovery and precision outcomes further confirmed its reliability and repeatability. Overall, the validated procedure is suitable for routine quality control and regulatory assessment of NDBA impurity in Simvastatin and Ezetimibe formulations, ensuring compliance with safety and quality standards.

7. REFERENCES:

1. Regulatory updates and analytical methodologies for nitrosamine impurities detection in sartans, ranitidine, nizatidine, and metformin along with sample preparation techniques. DOI: 10.1080/10408347.2020.1788375

2. A new analytical LC-MS/MS method for determination of eight standard nitrosamines (NDMA, NMBA, NDEA, NEIPA, NDIPA, NMPA, NDPA, NDBA) in a commercial small molecule drug product capsule and its active pharmaeceutical ingredient for treatment of fabry: a rare disease. DOI: org/10.1007/s11095-025-03875-7

3. Validation of analytical procedures Q2(R2)

4. Formation of n-nitrosamine drug substance related impurities in medicines: a regulatory perspective on risk factors and mitigation strategies. DOI: org/10.1021/acs.oprd.3c00153

5. Lipid-lowering effects of ezetimibe and simvastatin in combination. DOI: org/10.1586/erc.10.179

6. Structural mechanism for statin inhibition of HMG-CoA reductase. DOI: 10.1126/science.1059344

7. Ezetimibe therapy: mechanism of action and clinical update. DOI: org/10.2147/VHRM.S33664

8. P14-19 genotoxicity evaluation of n-nitrosamines: Focus on NDEA, NDBA, and NDBzA. DOI: https://doi.org/10.1016/j.toxlet.2025.07.397

9. Insight into the formation of n-nitrosodimethylamine in metformin products. DOI: 10.1016/j.jpba.2020.113877

10. A multi-analyte LC-MS/MS method for screening and quantification of nitrosamines in sartans. DOI: 10.38212/2224-6614.1063

11. Method for trace determination of N-nitrosamines impurities in metronidazole benzoate using high-performance liquid chromatography coupled with atmospheric-pressure chemical ionization tandem mass spectrometry. DOI: 10.1002/jssc.202200225

12. Development, optimization, and use of an apci source with temperature-controlled vaporization of solid and liquid samples. DOI: org/10.1007/s00216-012-6531-4

13. Analytical Methodologies to Detect N-Nitrosamine Impurities in Active Pharmaceutical Ingredients, Drug Products and Other Matrices. DOI: org/10.1021/acs.chemrestox.4c00234

14. Analytical method development and validation for determination of nitrosamine impurities n nitrosodibutylamine (NDBA) & nitroso of n-dimethyl erythromycin in sertraline HCl drug substance by LC-MS/MS. DOI:10.62441/nano-ntp.vi.2776

15. Liquid chromatography-mass spectrometry technique-a review DOI: 10.5958/0974-360x.2020.00097.9

16. Ultra-Sensitive Quantification of Tirzepatide in Human Plasma Using Xevo TQ Absolute Mass Spectrometry with waters_connect Quantitation Software.

17. Nitrosamine impurities in pharmaceuticals: an overview of regulatory requirements, risk assessment, and control strategies. DOI: 10.1016/j.xphs.2023.03.018

18. Nitrosamine drug substance-related impurities (NDSRIs): opportunities for efficient risk assessment and control. DOI: 10.1021/acs.oprd.2c00236

19. Lessons learned from the presence of N-nitrosamine impurities in medicines. DOI: 10.1016/j.xphs.2022.10.034

20. Analytical approaches for determination of nitrosamine impurities in pharmaceuticals using LC-MS/MS and high-resolution mass spectrometry. DOI: 10.1016/j.jpba.2021.114000

21. World Health Organization. Information note: Nitrosamine impurities in medicinal products. Geneva: World Health Organization; 2023. Nitrosamines are a class of highly potent genotoxic impurities

22. Potential for nitrosamine formation during pharmaceutical manufacturing processes and strategies for mitigation. DOI: 10.1021/acs.oprd.9b00355

Received on 19.05.2026 Revised on 03.06.2026

Accepted on 16.06.2026 Published on 10.07.2026

Available online from July 25, 2026

Asian Journal of Pharmaceutical Analysis. 2026; 16(3):178-184.

DOI: 10.52711/2231-5675.2026.00027

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License. Creative Commons License.

Trial	Column	Run Time (min)	Gradient Program (Time/%A: %B)	Modification Objective	Observation
Trial 1	Zorbax SB-Phenyl (150 × 4.6mm, 3.5 µm)	20	0.0 (80:20), 1.5 (80:20), 7.0 (50:50), 7.5 (5:95), 14.0 (5:95), 14.1 (80:20), 20.0 (80:20)	Initial chromatographic conditions	NDBA impurity and ezetimibe co-eluted at ~11.8 min
Trial 2	Zorbax SB-Phenyl (150 × 4.6mm, 3.5 µm)	30	0.0 (80:20), 1.5 (80:20), 7.0 (50:50), 14.0 (30:70), 20.0 (10:90), 25.0 (10:90), 25.1 (80:20), 30.0 (80:20)	Gradient program modified to improve separation	Improved NDBA-drug separation, but with longer run time, Simvastatin (~22.9 min)
Trial 3	Acclaim™ 120 Å (150 × 4.6mm, 3 µm)	20	0.0 (50:50), 1.5 (50:50), 10.0 (20:80), 11.0 (10:90), 14.0 (10:90), 14.1 (50:90), 20.0 (50:50)	Evaluation of different column chemistry	Impurity separated, but simvastatin did not fully elute within the run time
Trial 4	Zorbax SB-Phenyl (150 × 4.6mm, 3.5 µm)	26	0.0 (50:50), 1.5 (50:50), 10.0 (30:70), 15.0 (20:80), 15.1 (10:90), 20.0 (10:90), 20.1 (50:50), 26.0 (50:50)	Gradient re-optimization using phenyl column	NDBA (~11.7 min), ezetimibe (~12.3 min) and simvastatin (~17.2 min) were well separated