New Approaches for Simultaneous API and Counterion Analysis Using Charged Aerosol Detection

New Approaches for Simultaneous API and Counterion Analysis Using Charged Aerosol Detection Christopher Crafts1, Marc Plante1, Bruce Bailey1, Paul Gamache1, John Waraska1, Ian Acworth1, and Kannan Srinvasan2

1ESA−A Dionex Company, Chelmsford, MA, USA; 2Dionex Corporation, Sunnyvale, CA, USA

Christopher Crafts1, Marc Plante1, Bruce Bailey1, Paul Gamache1, John Waraska1, Ian Acworth1, and Kannan Srinvasan2

1ESA−A Dionex Company, Chelmsford, MA, USA; 2Dionex Corporation, Sunnyvale, CA, USA

AbstrActApproximately 50% of all drug molecules used in medicinal therapy are administered as salts. Salt forms are used to improve the biological and physiochemical properties of active pharmaceutical ingredients (APIs) including: aqueous solubility, hygroscopicity, solution pH, melting point, dissolution rate, chemical stability, crystal form, and mechanical prop-erties. Traditional approaches for salt analysis use several analytical methods to measure the API and various counterions. Recent innovations in mixed-mode column technology and universal detection techniques enable the rapid, sensitive, and simultaneous analysis of an API to-gether with its inorganic or organic counterion. The work in this study uses HPLC and the Corona® ultra™ detector for the analysis of several common pharmaceutical salts. A Dionex Acclaim® Trinity™ P1 column, which combines anion-exchange, cation-exchange, and reversed-phase functionalities, allows separation of positive and negative counterions, as well as lipophilic APIs in a single run. Different methods are presented, including an isocratic approach capable of simultaneously measuring the API and its counterion in less than 3 min. Figures of merit are reviewed, including accuracy, precision, linear range, LOD, and LOQ. Comparison of experimental and theoretical values are within 1%. The ability to test for ionic impurities of < 0.1% is demonstrated. The combination of the Trinity column and the Corona ultra detector offers a new, versatile, and powerful tool for counterion analysis with applications for salt selection, drug formulation, and drug impurity analysis.

IntroductIon Counterions influence an API’s solubility, stability, and hygroscopicity, so appropriate counterion selection is an important part of drug development process. The most common pharmaceutical counterions include: chloride, sodium, sulfate, acetate, phosphate, potassium, maleate, calcium, citrate, bromide, nitrate, ammonium, tosylate, phosphate, tartrate, and ethylenedi-amine.1,2 Methods for the analysis of 13 of the 16 most common pharma-ceutical salts are demonstrated on the zwitterionic or the nanopolymer silica

hybrid column technologies. The SeQuant polymeric zwitterionic column (ZIC®-pHILIC) uses zwitterionic stationary phase on porous polymer par-ticles. The separation is achieved by a hydrophilic interaction mechanism superimposed on weak electrostatic interactions.3 The Acclaim Trinity P1 column uses silica particles where the inner-pore area is modified with an organic layer that provides both reversed-phase and anion-exchange properties. The outer surface is modified with cation-exchange function-ality. The Charged Aerosol Detector (CAD®) uses a patented detection technique described in Figure 1 which is able to achieve near uniform re-sponse over a broad spectrum of analytes. The analysis of inorganic ions involves the use of volatile buffers, such as ammonium acetate or formate, to create salt particles that can then be charged and detected.

26802

1

2

3

4

5

6

7

89

101

75

6

10

98

4

3

2

6. Gas stream passes over corona needle

7. Charged gas collides with particles and transfer charge

8. High mobility species are removed

9. Remaining charged particles measured

10. Signal transferred to chromatographic software

1. Liquid eluent enters from HPLC system

2. Pneumatic nebulization occurs

3. Small particle droplets enter drying tube

4. Large droplets exit to drain

5. Dried particles enter mixing chamber

Figure 1. Corona ultra detector workflow.

2 New Approaches for Simultaneous API and Counterion Analysis Using Charged Aerosol Detection

sImultAneous sePArAtIon oF cAtIons And AnIons

Figure 3. Gradient analysis on the SeQuant ZIC-pHILIC, 4.6 × 150 mm, 5 µm column.

27130

2

22

Minutes

pA

0 2 4 6 8 10 12 14 16

Na+ K+

NO3– Cl–

Br–

Column: Acclaim Trinity P1 3.0 × 100 mm, 3 µmMobile Phase: Acetonitrile / 20 mM (total) Ammonium acetate pH 5.2 (60:40)Flow Rate: 0.5 mL/minInj. Volume: 5 µLColumn Temp.: 30 °C Sample Conc.: ~100 µg/mL eachDetection: Corona ultra 35 psi nitrogenFilter: Corona

27131

0.25

1.75

Minutes

pA ×

(100

,000

)

105 15 20 25

NO3–

Cl–

SO42–

Li+ Na+ K+

ClO4–

PO43–

Column: SeQuant ZIC-pHILIC 5 µm, 4.6 × 150 mm, 5 µmMobile Phase: A) Acetonitrile, isopropanol, 100 mM ammonium acetate, pH = 4.68, methanol (60:20:15:5) B) 30 mM Ammonium acetate pH = 4.68, acetonitrile, isopropanol, methanol (50:25:20:5)Flow Rate: 0.5 mL/minInj. Volume: 10 µLGradient: Time (min) 0 3 24 26 28 29 32 %B 20 20 70 70 15 20 20Column Temp.: 30 °CSample Conc.: ~30 µg/mL eachDetection: Corona ultra 35 psi NitrogenFilter: Corona

Figure 2. Isocratic analysis on the Acclaim Trinity P1, 3.0 × 100 mm, 3 µm column.

27132

0

400

Minutes

mV

30 6 9 12 15

1

2

3

45

6 78

9

10 1112

1314

15

16

Peaks: 1. Procaine 2. Choline 3. Tromethamine 4. Sodium 5. Potassium 6. Meglumine 7. Mesylate 8. Maleate 9. Chloride10. Bromide11. Iodide12. Phosphate13. Malate14. Tartrate15. Citrate16. Sulfate

Column: Acclaim Trinity P1, 3.0 × 50 mm, 3 µmMobile Phase: A) Acetonitrile B) Deionized water C) 200 mM Ammonium acetate pH = 4Flow Rate: 0.5 mL/minInj. Volume: 10 µLGradient: Time (min) 0 2 7 15 %B 35 35 0 0 %C 5 5 90 90Column Temp.: 30 °CSample Conc.: Between 35 and 150 µg/mL eachFilter: None

Figure 4. Gradient analysis on the Acclaim Trinity P1, 3.0 × 50 mm, 3 µm column.

reProducIbIlIty And rAnge

table 1. Intra- and between-day reproducibility of Five Analytes

% RSD Day 1 (n = 5)

% RSD Day 2 (n = 5)

% RSD Day 4 (n = 6)

% RSD Day 7 (n = 6)

% RSD All Days (n = 22)

Nitrate 1.25 1.45 1.00 0.88 2.67

Chloride 0.45 1.33 1.23 0.56 2.55

Sulfate 1.71 0.77 2.21 1.31 4.68

Sodium 1.94 1.35 1.69 1.54 3.30

Potassium 1.16 1.93 1.19 1.39 3.47

Percentage RSD of area averages of anions and cations using the method described in Figure 3.

3New Approaches for Simultaneous API and Counterion Analysis Using Charged Aerosol Detection

Figure 6. Diclofenac sodium salt, 1.3 µg on column.

27133

0

12,000

Area (pA * Minutes)

Amou

nt (n

g on

colu

mn)

2 4 6 8 10 12 14 160

Sodium Response Curve

Figure 5. Sodium acetate from 5–10,000 ng on column. Three injections, at each of the 12 concentrations, fit with a polynomial inversed x and y axis (r2 = 0.9997) overlaid with the 95% confidence limits.

API And counterIonsThe simultaneous measurement of an API and its counterion was il-lustrated using diclofenac sodium salt. Figure 6 shows the analysis on the Acclaim Trinity P1 column while Figure 7 shows the analysis on the ZIC-pHILIC column. In this example, the API and counterion elution order was reversed for the two columns with diclofenac retaining better on the Trinity column with a shorter run time.

This approach can be used to measure other APIs with counterions shown in Figures 2–4 on both the Trinity and ZIC columns.

27134

–5

80

Minutes1 2 3 4 5 60

Column: Acclaim Trinity P1, 3.0 × 50 mm, 3 µmMobile Phase: Acetonitrile/120 mM ammonium acetate pH = 4.8 (30 mM total buffer strength) (75:25) Flow Rate: 0.8 mL/minPeaks: 1. Sodium 2. Diclofenac

2

1

pA

27135

0

500

Minutes5 10 15 20 250

Column: ZIC-pHILICMobile Phase: A) Acetonitrile, isopropanol,100 mM ammonium acetate, pH = 4.68, methanol (60:20:15:5) B) 30 mM Ammonium acetate pH = 4.68, acetonitrile, isopropanol, methanol (50:25:20:5)Flow Rate: 0.5 mL/minInj. Volume: 10 µLGradient: Time (min) 0 3 24 26 28 29 32 %B 20 20 70 70 15 20 20Sample: Diclofenac sodium salt, 0.3 µgPeaks: 1. Diclofenac 2. Sodium

mV

1

2

Figure 7. Chromatogram of a 10 µL injection of diclofenac sodium salt with the method described in Figure 3.

FAst counterIon AnAlysIs

Figure 8. Naproxen sodium salt, ~500 ng on column.

27136

–60

80

Minutes

0

0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.00.0

12

12

pH = 3.8

pH = 5.2

Column: Acclaim Trinity P1, 3.0 × 50 mm, 3 µm Flow Rate: 1.0 mL/min Mobile Phase: Acetonitrile/120 mM ammonium acetate (30 mM total buffer strength) (75:25) pH varied as shown in figurePeaks: 1. Naproxen 2. Sodium

PA

lImIts oF detectIon And 0.1% ImPurItIes

table 2. lods for sodium and chloride using the corona ultra detector

ZIc-pHIlIc trinity P1

Sodium 1.3 ng 0.9 ng

Chloride 1.3 ng 0.2 ng

Limits of detection using the methods described in Figures 2 and 3. Limit defined as signal-to-noise > 3.

4 New Approaches for Simultaneous API and Counterion Analysis Using Charged Aerosol Detection

Figure 9. Three injections of diclofenac sodium salt, 5.7 µg on column.

Average of the chloride raw area values and the corresponding deviation along with the calculated weight percent of chloride in dicolfenac sodium samples run as described in Figure 9.

table 3. Area Values and deviation for chloride in diclofenac

sample raw Area (pA *min)

rsd (n = 5)

calculated chloride (w/w)

Diclofenac Standard 0.0231 13.01% 0.03%

Diclofenac + 0.1% Cl– 0.1069 2.39 0.14%

Diclofenac + 0.15% Cl– 0.1318 1.49% 0.17%

27137

–5.0

10.5

Minutes

0.0

0.01 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00

31

2

Column: Acclaim Trinity P1, 3.0 × 50 mm, 3 µm Mobile Phase: Acetonitrile/120 mM ammonium acetate (75:25) pH = 4.8 (30 mM total buffer strength)Flow Rate: 0.8 mL/min

Samples: A) Diclofenac sodium unspiked B) Spiked with 0.1% (w/w) chloride C) Spiked with 0.15% (w/w) chloridePeaks: 1. Sodium 2. Chloride 3. Diclofenac

pA C

B

A

Figure 10. Chromatogram of a 10 µL injection of diclofenac sodium salt 0.3 µg in acetonitrile/water (80:20) with 0.1% w/w of chloride using ZIC-pHILIC column and the method described in Figure 3.

27138

10

40

Minutes

mV

0 5 10 15 20 25

Column: ZIC-pHILIC Mobile Phase: A) Acetonitrile, isopropanol,100 mM ammonium acetate, pH = 4.68, methanol (60:20:15:5) B) 30 mM Ammonium acetate pH = 4.68, acetonitrile, isopropanol, methanol (50:25:20:5)Flow Rate: 0.5 mL/minInj. Volume: 10 µLGradient: Time (min) 0 3 24 26 28 29 32 %B 20 20 70 70 15 20 20Peaks: 1. Diclofenac 2. Chloride 3. Sodium

3

2

1

dIscussIonThe two columns evaluated (ZIC-pHILIC and Trinity P1) both provide the ability to simultaneously resolve anions and cations as shown in Figures 2–4. These columns use very different packing materials for their separations, which lead to different advantages depending on application requirements.

The ZIC-pHILIC column was found to operate best at a flow rate of 0.5 mL/min with a gradient method. It is possible to run the column isocratically at flow rates up 1 mL/min. However, long-term stability was compromised with a loss in peak shape and resolution. This was possibly due to build-up of trace ionic materials which do not fully elute under weak isocratic conditions.

Both columns were able to provide good reproducibility data (e.g., Table 1, ZIC data; Trinity data not shown) with large dynamic ranges (e.g., Figure 5, Trinity data; ZIC data not shown).

The focus of the work presented here was the analysis of an API and counterion on a single platform. This can be achieved using either columns as demonstrated in Figures 6–7. In these examples, compound specificity is reversed for the different columns. The API (diclofenac) is more retained on the Trinity column and the sodium peak elutes earlier. However, on the ZIC-pHILIC column, the diclofenac elutes very quickly, close to the void, and the sodium is more retained. Because the ZIC-pHILIC column uses a HILIC approach, the initial mobile phase composition needs to contain a high level of organic content, generally > 70%. For APIs that do not contain strong polar groups and are retained primarily by reversed-phase properties, the high organic mobile phase can lead to peaks eluting in or close to the void volume.This makes the analysis of hydrophobic APIs on the ZIC column difficult.

Figure 8 illustrates the seperation of another common API, naproxen, and its counterion sodium using the Trinity column. Baseline resolution can be obtained in less then 1 min. A simple change in the pH from 3.8 to 5.2 determines the elution order of the two major components. One of the major advantages found with the Trinity column was its ability to run fast isocratic methods. This resulted in better peak shape and sharper peaks which leads to lower limits of detection on the Corona ultra, as shown in Table 2.

Using the improved chromatography discussed here, an experiment was conducted to determine whether a significant difference between a diclofenac sodium sample spiked with 0.1% versus a sample spiked with 0.15% (w/w) chloride could be identified. Figure 9 is an overlay of chromatograms at each corresponding level. As shown in Table 3, the calculation of the chloride content with a four-point linear regres-sion curve from 3 to 25 ng chloride on column yielded recoveries of 100 ± 10% of the expected values. The method was sensitive enough to measure a trace amount of chloride contamination in the injection of the unspiked standard.

5New Approaches for Simultaneous API and Counterion Analysis Using Charged Aerosol Detection

Trinity and ultra are trademarks and Acclaim, CAD, and Corona are registered trademarks of Dionex Corporation.ZIC is a registered trademark of Merck SeQuant.

north America

U.S./Canada (847) 295-7500 south America

Brazil (55) 11 3731 5140

europe

Austria (43) 1 616 51 25 Benelux (31) 20 683 9768 (32) 3 353 4294 Denmark (45) 36 36 90 90 France (33) 1 39 30 01 10 Germany (49) 6126 991 0 Ireland (353) 1 644 0064 Italy (39) 02 51 62 1267 Sweden (46) 8 473 3380 Switzerland (41) 62 205 9966 United Kingdom (44) 1276 691722

Asia Pacific

Australia (61) 2 9420 5233 China (852) 2428 3282 India (91) 22 2764 2735 Japan (81) 6 6885 1213 Korea (82) 2 2653 2580 Singapore (65) 6289 1190Taiwan (886) 2 8751 6655

dionex corporation

1228 Titan Way P.O. Box 3603 Sunnyvale, CA 94088-3603 (408) 737-0700 www.dionex.com

LPN 2610-01 10/10©2010 Dionex Corporation

conclusIon • AnalysisusingCADandeithertheAcclaimTrinityP1orthe

SeQuant ZIC-pHILIC column provided sensitive and reproducible results for the determination of API counterions.

• TheHILICchemistryandcolumnpackingmaterialforthe ZIC-pHILIC column made it necessary to use longer gradient runs to achieve optimal performance. The need to run under HILIC conditions make it unsuitable for analyzing hydrophobic APIs.

• Themixed-mode,NanopolymerSilicaHybrid(NSD)technologyemployed in the Trinity P1 column allowed for fast isocratic API counterion analysis required when screening counterions and in QC functions.

reFerences1. Kumar, L.; Amin, A.; Bansal, A.K. Salt Selection in Drug Develop-

ment, Pharm Tech. 2008,3 (32), 128–146. 2. Haynes, D.A.; Jones, W., Samuel Motherwell, W.D. Occurrence of

Pharmaceutically Acceptable Anions and Cations in the Cambridge Structural Database. J. Pharm. Sci. 2005, 94, 2111–2120.

3. Merck website on ZIC-HILIC column http://www.sequant.com.

Acknowledgements ESA is grateful to Dr. Xiaodong Liu (Dionex) for providing the data on gradient analysis using the Trinity P1 column.