Impurity Profiling with the Agilent 1200Series LC System Part 4: Method Validation of a Fast LCMethod

Abstract

Analytical laboratories working in a regulated environment have to vali-

date their methods, to ensure that results fulfill all regulatory require-

ments. The validation procedure introduced in this Application Note was

based on recommendations from the U.S. Pharmacopeia (USP) and the

ICH guidelines Q2B respectively. A fast LC method for one main com-

pound and its four impurities is successfully validated.

A. G. Huesgen

Application Note

min0 0.5 1 1.5 2 2.5

mAU

0

0.5

1

1.5

2

2.5

0.05 % level

0.027 %level = LOQ

0.007 %level = LOD

IntroductionAnalytical laboratories working ina regulated environment must val-idate their methods in order toensure that the results fulfill allregulatory requirements. In addi-tion, the results from differentusers in different laboratories arecomparable, even though separateequipment was used. Consequently,the method is required to be asrobust as possible to compensatefor variations, which might occurif different users perform thesame analysis on the same or dif-ferent equipment. The validationprocedure introduced in thisApplication Note was based onrecommendations from the U.S.Pharmacopeia (USP), ICH guide-lines Q2B1,2, as well as FDA guide-lines4,5, which are recognizedworldwide and employed by ana-lytical laboratories in the pharma,food, environmental and chemicalindustry. In the following experi-ment, a fast LC method for onemain compound and its impuritiesis validated, (see figure 1 for asample chromatogram). Methodparameters were obtained fromthe method development group(reference 3).

ExperimentalAn Agilent 1200 Series RapidResolution LC system was usedwith the following modules:• Agilent 1200 Series binary pump

SL and vacuum degasser • Agilent 1200 Series high-perfor-

mance autosampler SL • Agilent 1200 Series thermostat-

ted column compartment SL

• Agilent 1200 Series diode-arraydetector SL

• Data acquisition and evaluationsoftware: Agilent ChemStationB.02.01.SR1

• ZORBAX SB C-18 RRHTcolumns with internal diametersof 4.6 mm and lengths of 50 mm,packed with 1.8-µm particles

• Main compound and pure impu-rities were obtained using thepurification procedure decribedin reference 6.

Validation procedureHaving done some pre-validationexperiments, the following valida-tion protocol for the above-described method3 was set up:

Specificity was tested and is given,see reference 3. No further samplepreparation steps were taken. Thesample compounds were weighedand dissolved in water.

Results and discussionValidation of main compound

Sample preparation:The stock solution of the maincompound contained 2.3 mg/mL.This solution was diluted to givethe following desired concentra-tions:

Main compound:• Stock solution: 2.3 mg/mL (used

for carry over and linearity tests)

2

min0.5 0.75 1 1.25 1.5 1.75 2 2.25 2.5

mAU

-0.2

0

0.2

0.4

0.6

0.8

1

0.8

69 -

D

1.4

14- A

1.5

46 -

Mai

n co

mpo

und

2.4

90 -

C

2.6

12 -

B

Figure 1Analysis of the main compound and its 4 impurities. Main compound 1.15 mg/mL, impurities 0.072 % of the main compound.

Validation protocol :1. Precision of areas and RT of the main compound 6 concentrations, 6 runs each2. Accuracy of main compound 6 concentrations, 6 runs each3. Linearity of main compound 6 concentrations, 6 runs each4. Carry over for main compound 3 injections of stock solution 5. Range of main compound6. Precision of areas and RT of impurities 7 concentrations, 6 runs each7. Accuracy of impurities 7 concentrations, 6 runs each8. Linearity of impurities 7 concentrations, 6 runs each9. Range of impurities10. Limit of Detection and LOQ11. Robustness of main compound and impurities Different column temperatures, flow rates,

injection volumes, TFA concentrations, gradient steepness, wavelength, users and instruments, no ruggedness tests

• Dilution 1: 1.15 mg/mL (used forprecision, accuracy, linearity androbustness tests)

• Dilution 2: 0.575 mg/mL (usedfor precision, accuracy, linearitytest)

• Dilution 3: 0.288 mg/mL (usedfor precision, accuracy, linearitytest)

• Dilution 4: 0.144 mg/mL (usedfor precision, accuracy, linearitytest)

• Dilution 5: 0.072 mg/mL (usedfor precision, accuracy, linearitytest)

1. Precision of retention times andareas The results for retention time andarea precision for all different con-centration levels are summarizedin figure 2. The precision limit forretention times is 0.1 % rsd. Theprecision limit for areas is 2 % rsd.For all concentrations the limitsfor retention time and area preci-sion are fulfilled. To achieve suffi-cient resolution from the impuri-ties, the concentration of the maincompound should be <1.2 mg/mL.

2. Accuracy of main compound The accuracy was tested using theabove-mentioned concentrations.A maximum deviation of 2 % wasset as the limit. All concentrationspassed the requirement, (figure 3).

3. Linearity for main compoundThe linearity was tested using all 6concentrations. A correlation coef-ficient of > 0.99990 was set as thelimit for this concentration range.The determined correlation coeffi-cient was 0.99999. The responsefactors are within the 5 % limitfrom 2.3 down to 0.073 mg/mL(figure 4).

0

0 .2

0 .4

0 .6

0 .8

1

2 .3 1 .1 5 0 .5 8 0 .2 9 0 .1 4 0 .0 70

0 .2

0 .4

0 .6

0 .8

1

2 .3 1 .1 5 0 .5 8 0 .2 9 0 .1 4 0 .0 7

Rsd RT for 6 runs of each concentration

rsd

RT

rsd

area

0

0 .2

0 .4

0 .6

0 .8

1

2 .3 1 .1 5 0 .5 8 0 .2 9 0 .1 4 0 .0 7

0

0 .1

0 .2

0 .3

0 .4

0 .5

0

0 .1

0 .2

0 .3

0 .4

0 .5

Concentration (mg/mL)

Concentration (mg/mL)

2 .3 1 .1 5 0 .5 8 0 .2 9 0 .1 4 0 .0 72 .3 1 .1 5 0 .5 8 0 .2 9 0 .1 4 0 .0 72 .3 1 .1 5 0 .5 8 0 .2 9 0 .1 4 0 .0 7

Rsd of areas for 6 runs of each concentration

rsd RT for 6 runs each concentration

rsd of areas for 6 runs each concentration

Figure 2Precision of retention times and areas of different concentrations of the main compound; 6 runsfor each concentration.

Accuracy of different concentrations

-1

-0.5

0

0.5

1

1.5

2

0 0.5 1 1.5 2 2.5

Concentration (mg/mL)

Dev

iatio

n in

%

deviation in %from spiked value

Figure 3Accuracy for different concentrations of the main compound, 6 runs for each concentration.

Linearity of the main compound from 0.72 to 2.3 mg/mL

0

0.5

1

1.5

72 144 288 575 1150 2300Amount

Amou

nt/A

rea

Response factors

within 5% limit

Figure 4Linearity of the main compound.

3

4. Carry-over of main compoundThe carry-over was evaluated byinjecting the stock solution 6 timesfollowed by the injection of 5 µLpure water. The carry-over wasfound to be ~ 0.01 %.

5. Range of main compoundThe range for the main compoundwith good precision, accuracy andlinearity lies between 2.3 down to0.073 mg/mL.

Validation of impurities

A, B, C, and D

Sample preparationThe impurities were analyzed bypreparing different concentrationlevels using the pure impuritycompounds. The stock solutioncontained 4.9 mg/mL for impurityA, C, and D and 4.6 mg/mL of com-pound B. This solution was dilutedby a factor of 1:1000 to obtain aconcentration in the µg/mL range.The performance was evaluatedfor 7 impurity concentrationsbased on the diluted mixtures, seetable 1.

6. Precision of retention times andareas of impuritiesPrecision of areas was evaluatedfor 6 concentrations from 0.153 upto 4.9 µg/mL for compounds A, Cand D, and from 0.144 to 4.6 µg/mLfor compound B. Precision ofretention times was evaluated forall 7 concentrations. A summary ofall results is shown in figure 5. The0.05 % level fulfilled the acceptablelimit of < 5 % rsd for areas. The0.027 % level showed an area pre-cision < 6 % rsd, which is withinthe 10 % limit.The 0.013 % levelshowed an area precision of < 14 %,which is acceptable for this lowconcentration. Retention time pre-cision is below 0.5 % rsd for all

Figure 5RSD of retention times and areas for impurities; 6 runs for each concentration .

Precision of RT of impurities

Precision of areas of impurities

0

24

68

10

1214

16

0 2 4 6

rsd

Area

(%)

0.013 % level

0.025 % level

0.05 % level

0.1 % level

0.2 % level

0.4 % level

0

0.05

0.1

0.15

0.2

0.25

0 2 4 6

Concentration (μmg/mL)

Concentration (μg/mL)

0.007 % level

0.013 % level

0.025 % level

0.05 % level

0.1 % level

0.25 % level

0.43 % level

4

Table 1Dilution series for impurities.

Impurity Dilution Dilution Dilution Dilution Dilution Dilution Stock solution6 µg/mL 5 µg/mL 4 µg/mL 3 µg/mL 2 µg/mL 1 µg/mL µg/mL(% level)* (% level)* (% level)* (% level)* (% level)* (% level)* (% level)*

A 0.077 0.153 0.306 0.613 1.225 2.45 4.9 (0.007% (0.013%) (0.027%) (0.05%) (0.107%) (0.213%) (0.426%)

B 0.072 0.144 0.287 0.575 1.150 2.30 4.6 (0.006%) (0.013%) (0.025%) (0.05%) (0.1%) (0.2%) (0.4%)

C 0.077 0.153 0.306 0.613 1.225 2.45 4.9(0.007% (0.013%) (0.027%) (0.05%) (0.107%) (0.213%) (0.426%)

D 0.077 0.153 0.306 0.613 1.225 2.45 4.9(0.007% (0.013%) (0.027%) (0.05%) (0.107%) (0.213%) (0.426%)

* Percentage is based on a main compound concentration of 1.15 mg/mL

concentrations and passed the setlimit of rsd < 0.5 %. An exampleof the precision of retention timesand areas is shown in figure 6. Tenchromatograms from 10 consecu-tive runs were superimposed andthe precision of retention times isfor all compounds < 0.16 %.

7. Accuracy of impuritiesAccuracy of impurities was evalu-ated for the 0.027 %, 0.05 %, 0.107% and 0.213 % level of impuritiesA, C and D. The 0.025 %, 0.05 %,0.1 % and 0.2 % level of impurity Bwas evaluated. Spiked sampleswith known concentrations wereanalyzed and data were evaluatedusing auto-integration and the cali-bration parameters used for preci-sion measurements. The deviationfrom the spiked value should notbe more than ± 5 %. The maximumdeviation is ± 4.3 % for the deter-mined concentration ranges (fig-ure 7). Each concentration wasinjected 6 times and the averagevalue was used as the calculatedamount.

8. Linearity of impuritiesThe linearity of all impurities wastested using all 7 concentrationlevels. A correlation coefficient of> 0.9990 was set as the limit forthis concentration range. Theestablished correlation was0.9998. Linearity based on theresponse factors is calculatedbetween 4.9 µg/mL (impurities A,C, and D), 4.6 µg/mL (impurity B)down to 0.306 µg/mL (impuritiesA, C, and D), and 0.287 µg/mL(impurity B); an example is givenin figure 8. The response is withinthe 5 % limit for these concentra-tion ranges.

Figure 7Determination of accuracy of impurities A, B, C, and D. 6 runs for each concentration with maximum deviation ± 4.5 % for the 0.027 % level.

Accuracy of impurities

1 2 3

Spiked amount (g/mL)

Devi

atio

n in

%

0.027 % impurity level

0.05 % impurity level

0.213 % impurity level

0.1 % impurity level0

-6-8-10

-4-2

246810

Figure 8Linearity of impurity B across a concentration range between 0.077 to 4.9 µg/mL.

Linearity of impurity B from 0.072 to 4.6 mg/mL

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

0.07 0.14 0.29 0.58 1.15 2.3 4.6Amount (μg/mL)

Amou

nt/a

rea

Responsefactors

5% rangeLinear range

Figure 6Overlay of 10 chromatograms within 1 sequence, RSD RT for all peaks < 0.16 %.

Rsd RT <0.16%

min0.5 0.75 1 1.25 1.5 1.75 2 2.25 2.5

mAU

-0.4

-0.2

0

0.2

0.4

0.6

0.8

5

9. Range of impuritiesThe range of impurities with an acceptable precision, accuracyand linearity is between 0.306 µg/mL (0.287 µg/mL impurity B) and 4.9 µg/mL (4.6 µg/mL impurity B).

10. Limit of detection (LOD) and limitof quantitation (LOQ) for impuritiesThe limit of detection was deter-mined using the 0.007 % level ofcompounds A, C and D. For com-pound B the 0.006% level wasused. In figure 9 the resultingchromatogram, lower trace, isshown. All impurities at that con-centration level have a signal tonoise ratio S/N > 2 (table 2).In figure 9 chromatograms of 2further concentration levels forthe impurities related to the maincompound concentration of 1.15mg/mL are shown. The trace in themiddle shows the 0.027 % level(impurities A, C andD) and the0.025 % level (impurity B), whichis the limit of quantitation, and theupper trace shows the 0.05 %level. In table 2 the results forLOD and LOQ are summarized.The limit of quantitation was evaluated for the 0.027 and the0.025 % concentration levelsrespectively. The signal to noiseratio limit for the LOQ is 10 andthe values shown for the 0.027 and 0.025 % levels are proximate.The 0.05 % is clearly above the setlimit. The area precision for the0.05 % level is < 2.6 %, and for the0.027 and 0.025 %levels the areaprecision is < 5.4 %.

11. Robustness of method for maincompound and its impuritiesTo test the robustness of themethod, the main compound wasdissolved in water with a concen-

(figure 10). In table 3 the resultsfor the main compound and impurities are summarized. Theonly critical parameter is the

tration of 1.15 mg/mL. This solu-tion was spiked with impurities toachieve an impurity concentra-tion level of approximately 0.07 %

Figure 9Chromatogram of impurities close to the limit of detection at the 0.00 7 % level (lower trace). The trace in the middle shows the chromatogram at the 0.027 % level, close to the limit of quantitation. The upper trace shows a chromatogram of the 0.05 % level.

min0 0.5 1 1.5 2 2.5

mAU

0

0.5

1

1.5

2

2.5

0.05 % level

0.027 %level = LOQ

0.007 %level = LOD

Figure 10Chromatogram of sample used for robustness tests.

min0.5 0.75 1 1.25 1.5 1.75 2 2.25 2.5

mAU

-0.2

0

0.2

0.4

0.6

0.8

1

0.8

69 -

D

1.4

14- A

1.5

46 -

Mai

n co

mpo

und

2.4

90 -

C

2.6

12 -

B

6

Table 2LOD and LOQ for impurities.

Impurity concentration LOD S/N LOQ S/N

A 0.007 % level 2.8 0.027 %level 9.5B 0.006 % level 2.1 0.025 %level 9.1C 0.007 % level 3.8 0.027 % level 13.1D 0.007 % level 3.4 0.027 %level 10.1

wavelength. The wavelengthshould not vary more than 1 nm.All other parameter changes causedeviations for the areas of lessthan 2 %, which is acceptable for amain compound. The results forthe impurities are also shown intable 3. All results except thewavelength change are within the10 % limit for the areas. The wave-length should not vary more than1 nm.

In figure 11 an example is shownfor the day-to-day repeatability forretention times and areas. Theresults of 3 sequences are over-laid. Each sequence contained 10runs and were analyzed on 3 con-secutive days. The instrument wasturned off overnight.

Figure 11Day-to-day precision of retention times and areas, overlay of 3 sequences with 10 runs each.Repeatability of retention time day-to-day (figure 11a) < 0.6 %, repeatability for areas <4.5 %, difference between highest and lowest amount < 16 % for the amounts (figure 11b).

RT variation 1st day

RT variation 2nd day

RT variation 3rd day

Amounts 1st day

Amounts 2nd day

Amounts 3rd day

Retention time variation of impurity A over 3 days

1.35

1.37

1.39

1.41

1.43

1.45

1.47

1.49

0 5 10Run number

Rete

ntio

n tim

e (m

in)

Variation of amounts for compound A over 3 days

5.50E-01

6.00E-01

6.50E-01

7.00E-01

7.50E-01

8.00E-01

8.50E-01

9.00E-01

0 2 4 6 8 10Run number

Amou

nt (μ

g/m

L)

A

B

Table 3Robustness test results for main compound and impurities.

Parameter changed Deviation of amounts Resolution between Deviation of amountsfor the main compound main compound and (%) for impurities9%) impurity A (spiked amounts fig. 10)

Flow ± 2 % 1.96 % rsd 2.56 for + 2 % change < 9.8 % rsd for 2.55 for - 2 % change amounts

Column temperature 0.17 % rsd 2.61 for 5 % change± 5 % 2.51 for + 5 % change < 5 % rsd for amountsGradient slope ±10 % 0.07 % rsd 2.59 for + 10 %change < 3 % rsd for amounts

2.51 for - 10 % changeInjection volume 0.001 % from expected 2.45 for + 5 % change 6.5 % from expected ± 5 % amount 2.62 for - 5 % change amount TFA concentration 0.07 % rsd 2.49 for -10 % change± 10 % 2.61 for +.10 %change < 4 % rsd for amounts Wavelength ± 3 nm ± 10 % for area counts 2.61 ± 18 % for area countsDay-to-day 0.34 % rsd for 2.60-2.63 < 0.6 % for retentionrepeatability retention times timeDay-to-day 0.35 % rsd for amounts 2.60 – 2-63 < 4.5 % for amountsrepeatability and 1.15 mg/mL

weighted sample3 different instruments, ± 0.47 % deviation 2.6 -3.1 ± 10 % for 0.025 % different columns, for 1.15 mg/mL level for amountsintermediate precision ± 4.5 % for 0.05 %

level for amounts

7

© 2006 Agilent Technologies, Inc.

Published October 1, 2006Publication Number 5989-5620EN

www.agilent.com/chem/1200

Results of method validationIn table 4 the results of themethod validation are summa-rized. The set limits are fulfilled.Special attention is required forthe wavelength. The wavelengthvariation should not be more than± 1 nm. Typically, a wavelengthvariation of ± 3 nm is consideredacceptable. In this experiment, thelimits for wavelength variationsare more restrictive, based on theresults.

ConclusionA fast LC method was developedfor the analysis of a main com-pound and four impurities. Thevalidation of this method was suc-cessful. All requirements regardingprecision, linearity, accuracy androbustness were fulfilled. This sig-nifies that the fast LC method canbe used in QA/QC labs and is com-pliant with USP/ICH recommenda-tions. Faster LC methods providethe same data quality and, as anadditional benefit, higher samplethroughput.

References1. International Conference onHarmonization (ICH) Q2B.Validation of AnalyticalProcedures: Methodology; Nov.1996, published in the “FederalRegister”, Vol 62, No. 96, pages

27463-27467, May 19, 1997.

2. Internet Resources:http://fda.gov/cder/guidance/index.htm and http://www.labcompli-ance.com, Dr. Huber’s website

Anglika Gratzfeld-Huesgen is

Application Chemist at Agilent

Technologies, Waldbronn,

Germany.

Table 4Results of method validation.

Parameters

Precision for areas• Main compound < 2 % for all experiments passed• Impurities < 10 % at the limit of quantitation passed

Accuracy for main compound < ± 2 % passedAccuracy for impurities at LOQ ± 5 % passedPrecision of Retention times < 0.5 % within 1 series passedPrecision of retention time < 2 % day-to-day passedLinearity > 0.999 passedResolution > 2 for all peaks passedLOD S/N > 2 at the ~ 0.007 % level for impurities passedLOQ S/N > 10 at the ~ 0.027 % level for impurities passedRange of the main compound: 2.3 to 0.073 mg/mL passedRange of the impurities: 4.9 (0.426 % level) to 0.287 µg/mL (0.025 % level) passedRobustness tests for area deviation: < 2 % for main compound passed*Robustness tests for area deviation: < ± 5 % for impurities at the 0.05 %level Passed*

*wavelength variations of ± 1 nm are acceptable and should be carefully controlled

3. Michael Frank “ Impurity Profilingwith the Agilent 1200 LC SystemPart 3: Rapid Condition Scoutingfor Method Development”, Agilent

Application Note, publication

number 5989-5619EN, 2006.

4. Reviewer Guidance, Validation ofChromatographic Methods, Centerfor Drug Evaluation and Research,Food and Drug Administration,1994.

5. Guideline for Submitting Samplesand Analytical Data for MethodsValidation, Food and DrugAdministration, 1987.

6 Udo Huber “Impurity Profilingwith the Agilent 1200 Series LCSystem Part 2: Isolation ofImpurities with PreparativeHPLC”, Agilent Application Note,

publication number

5989-5618EN, 2006.