HPLC Method       HPLC Method       DevelopmentDevelopment

Part 1. General Chromatographic Theory

Part 2. The Stationary Phase: An Overview of HPLC Media 

Part 3. The Role of the Mobile Phase in Selectivity

Part 4. Column Care and Use

3

The Liquid Chromatographic Process

4Mikhail Tsvet, 1872-1919

Origin of Liquid Chromatography

Empty Column

Adsorbent particles added

Sample is loaded onto the top of

the column

Solvent is added to the

top of the column

Separation occurs as the bands

move down the column

Column Chromatography:

5

Basis of Chromatographic SeparationSeparation of compounds by HPLC depends on the interaction of analyte molecules with the

stationary phase

and the

mobile phase.

Mobile Phase Stationary Phase

6

Mobile Phase

Stationary Phase

Analytes

The Separation Process

7

Mobile Phase

Stationary Phase

Analytes

The Separation Process

8

Non-Polar Stationary Phase (e.g. C18)

O H

H

OH

H OH

H

N

N

N

N

N

O H

H

OH H

Polar/AqueousMobile Phase

Benzo(a)pyrene Ethyl sulfate

The Separation Process

9

Ethyl sulfateBenzo(a)pyrene

Hydrophobic Interactions

10

In any separation, almost never get a pure, single mode of separation. In RP, performance will be dictated by mixture of:

1. Hydrophobic interactions

2. Polar interactions

3. Ionic interactions

Method Development

= modulating stationary phase and mobile phase conditions to optimize these interactions and achieve a specific

separation goal.

OH

N

CH3

CH3

CH3

CH3

Tapentadol

Mechanisms of Interaction

11

S iO

S iO

S iO

S iO

S iO

S iO

S iO

S iO

S iO

S iO HCH 3

OH O H OH O H O H O H OH O H O H O H

O O O- O H O O OO

S iS i S i

S iS iCH 3

C H 3

CH 3C H 3

CH 3C H 3

CH 3 C H 3

C H 3C H 3 C H 3 CH 3

C H 3

C H 3

C H 3O H

OH

N

CH3

CH3

CH3CH3

Hydrophobic•

Weak, transient interactions between a non-polar stationary phase and the molecules

•

Hydrophobic & van Der

Waals interactions

•

Retention will be predicted by Log P

values

Hydrophobic Interactions

12

S iO

S iO

S iO

S iO

S iO

S iO

S iO

S iO

S iO

S iO HCH 3

OH O H OH O H O H O H OH O H O H O H

O O O- O H O O OO

S iS i S i

S iS iCH 3

C H 3

CH 3C H 3

CH 3C H 3

CH 3 C H 3

C H 3C H 3 C H 3 CH 3

C H 3

C H 3

C H 3O H

OH

N

CH3

CH3

CH3CH3

Polar

•

Interactions between polar functions groups of analyte and residual silanols or polar groups on media

•

Hydrogen bonding, dipole-

dipole interactions

• Relatively weak and transient

Polar Interactions

13

S iO

S iO

S iO

S iO

S iO

S iO

S iO

S iO

S iO

S iO HCH 3

OH O H OH O H O H O H OH O H O H O H

O O O- O H O O OO

S iS i S i

S iS iCH 3

C H 3

CH 3C H 3

CH 3C H 3

CH 3 C H 3

C H 3C H 3 C H 3 CH 3

C H 3

C H 3

C H 3O H

OH

NH+

CH3

CH3

CH3CH3

Ion-Exchange

•

Interactions between ionizable functional groups on analyte and counter-charged moiety on stationary phase

• Ion-exchange

• Strong, slow interactions

Ion‐Exchange Interactions

14

Chromatographic Measurements

15 k’ Asym N α

Chromatographic Measurements

16

Analytes which do not interact with the adsorbent elute from the

column in a volume equal to the void volume

in the column. The void volume of a column is the amount of mobile phase in the column between the adsorbent particles and in the pores of the porous adsorbent particles.

Mobile phase occupies the space between the particles or the interstitial volume.

Mobile phase fills the pores of the porous adsorbent particles.

The Void Volume

17

A compound which does not interact with the adsorbent at all elutes at what is termed the void volume

or the solvent front. The time that it takes for non-retained components to elute is the void time or t0

.

• void volume•

solvent front• t0

The Void Volume

18

The capacity factor

of the eluting compound is its elution volume (time) relative to the elution volume (time) of an unretained

compound. The k’

value for a given analytes will be determined by its relative affinity for the stationary phase and mobile phase.

t0

Capacity factor (k’)=

tR

– t0

t0

Capacity Factor (k’)

19

For any given analyte, the k’

value will be most readily modified by changing the % of strong mobile phase

(e.g. methanol or ACN).

Example: The Separation of Steroids:Column used: C18(2) 150 x 4.6 mm

Capacity Factor (k’)

20

Effect of Changing % Organic

21

The asymmetry

value (Asym) for a peak is a measure of the divergence of the peak from a perfect Gaussian peak shape. Peaks with asymmetry values greater than 1.0 are said to be “tailing”, those with asymmetry values less than 1.0 are said to be “fronting”.

Asymmetrical peaks can be attributed to a number of factors:(1) Strong secondary interactions

(e.g. ionic interactions between basic compounds and residual silanols)(2) Sample overloading(3) Sample solvent incompatibility(4) Poor packing

Peak Asymmetry

Classical peak tailing in reversed-phase methods is most commonly caused by strong ionic interactions

between basic analytes and residual silanols on the surface of the silica.

S iO

S iO

S iO

S iO

S iO

S iO

S iO

S iO

S iO

S iO HCH 3

OH O H OH O H O H O H OH O H O H O H

O O O- O H O O OO

S iS i S i

S iS iCH 3

C H 3

CH 3C H 3

CH 3C H 3

CH 3 C H 3

C H 3C H 3 C H 3 CH 3

C H 3

C H 3

C H 3O H

O H

NH+

CH 3

C H 3

CH 3CH 3

Ion-Exchange

22

Peak Tailing due to Secondary Interactions

2323

Example: The Separation of Tricyclic Antidepressants:Column used: Column A C18 50 x 2.1 mm

Column B C18 50 x 2.1mm

Amitriptyline (pKa 9.4) Nortriptyline (pKa 9.7) Protriptyline

(pKa 8.2)

Peak Tailing due to Secondary Interactions

24

Sample 28 (Halo C18-50x2, 2.7 um, MeOH:0.1% ... Max. 1.1e5 cps.

3.4 3.6 3.8 4.0 4.2 4.4 4.6 4.8 5.0 5.2 5.4 5.6 5.8 6.0Time, min

1

2 3

45

6

7 8

Sample 28 (Halo C18-50x2, 2.7 um, MeOH:0.1% ... Max. 1.1e5 cps.

3.4 3.6 3.8 4.0 4.2 4.4 4.6 4.8 5.0 5.2 5.4 5.6 5.8 6.0Time, min

4

6

Mobile phase: A = 0.1% Formic acid in water, B = 100% Methanol,Gradient: 15 to 60% B in 5 min, hold for 1 minFlow rate: 400 µL/min1. Amoxapine2. Imipramine3. Desipramine4. Protriptyline*5. Amitriptyline6. Nortriptyline*7. Clomipramine8. Norclomipramine

Sample 7 (TCA-Kin-XB-C18, 50x2, 2.6, MeOH 0.... Max. 1.2e5 cps.

3.4 3.6 3.8 4.0 4.2 4.4 4.6 4.8 5.0 5.2 5.4 5.6 5.8 6.0Time, min

2 3

1

4

5

6

78

15

Sample 13 (Kinetex XB-C18-50x2, 2.6 um, MeO... Max. 1.2e5 cps.

3.4 3.6 3.8 4.0 4.2 4.4 4.6 4.8 5.0 5.2 5.4 5.6 5.8 6.0Time, min

4

6

Column B

C18 50x2.1mmColumn A

C18 50x2.1mm

Peak tailing

Peak Tailing due to Secondary Interactions

25

Strongly basic analytes are very sensitive to sample loading, and will display peak tailing effects as a function of increasing loading (µg on column):

min12 12.5 13 13.5 14 14.5 15 15.5 16 16.5

mAU

0

20

40

60

80

100

DAD1 C, Sig=254,4 Ref=300,100 (DJ040611\DJGSK 2011-04-06 12-28-48\GSK000010.D)

14.15

7

12 µg on column; USP Tailing = 1.87

min12 12.5 13 13.5 14 14.5 15 15.5 16 16.5

mAU

0

10

20

30

40

50

DAD1 C, Sig=254,4 Ref=300,100 (DJ040611\DJGSK 2011-04-06 12-28-48\GSK000008.D)

14.16

9

5 µg on column; USP Tailing = 1.34

min12 12.5 13 13.5 14 14.5 15 15.5 16 16.5

mAU

0

5

10

15

20

25

30

DAD1 C, Sig=254,4 Ref=300,100 (DJ040611\DJGSK 2011-04-06 12-28-48\GSK000006.D)

14.20

8

2.5 µg on column;USP Tailing = 1.17

pKa 9.7

Peak Tailing due to Overloading

26

min1.8 1.9 2 2.1 2.2 2.3 2.4 2.5 2.6 2.7

mAU

0

250

500

750

1000

1250

1500

1750

2000

2250

DAD1 A, Sig=240,10 Ref=off (MT042211\DJGSK 2011-04-22 09-12-57\MUPIROCIN000001.D)

5

1.905

2.145

2.495

Fronting due to overload

Neutral and acidic compounds will typical show peak fronting

when the column is overloaded.

Detector saturation!

Peak Fronting due to Overload

27

Peak fronting

can also be due to sample

solvent effects:(1)

Sample insolubility(2)

Sample solvent is too strong:

0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.00.0

2.0e4

4.0e4

6.0e4

8.0e4

1.0e5

1.2e5

1.4e5

1.6e5

1.8e5

2.0e5

2.2e5

0.24

1.791.36

0.970.51 1.71 1.85 3.972.15 3.213.36

Sample in 100% Methanol

Breakthrough!

Fronting

0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.00.0

5000.0

1.0e4

1.5e4

2.0e4

2.5e4

3.0e4

3.5e4

4.0e4

4.5e4

5.0e4

5.5e4

6.0e4

6.5e4

7.0e4

1.061.77

Sample in 50% Methanol

Morphine

Hydromorphone

Norhydrocodone

Peak Fronting due to Solvent Effects

28

Selectivity

is a measure of the difference in the interactions of two compounds with the stationary phase. Selectivity is a function of both the stationary phase and the mobile phase.

α

= k2

/k1

Selectivity (α)

2929

The choice of stationary phase

will often have a dramatic effect on the selectivity of analytes.

OH

H H

OCH3

H

Estrone

OH

H H

OHCH3

H

Estradiol

m i n1 2 3 4 5

m A U

0

2 5

5 0

7 5

1 0 0

1 2 5

1 5 0

1 7 5 C18 Column 1+2

α

= 1.0

Phenyl Column 1

2

m i n1 2 3 4 5

m A U

0

2 0

4 0

6 0

8 0

1 0 0

α

= 2.3

Selectivity (α)

30

But mobile phase

is also a very powerful tool in modulating selectivity.

35% Methanol

20% Acetonitrile

Column:

Gemini 5 µm C6-Phenyl, 150 x 4.6mm

Mobile phase: 20mM KH2

PO4

, pH 2.5; % organic as noted

Flow rate:

1.0 mL/minDetection:

UV @ 220nm

1. Saccharin2. p-Hydroxybenzoic

Acid3. Sorbic

Acid4. Dehydroacetic

Acid5. Methylparaben

Selectivity (α)

3131

The sample is injected in a narrow band which then enters the bed of the column. The peaks become broader as a function of many variables, including:

• The size

of the particles• How well the particles are packed

into the column• Particle morphology

w1/2

tRInjected Sample Band

Column Efficiency (N)

32

W

W1/2

tR

Peak Width (W)

The amount of band (peak) broadening or dispersion that occurs in the column is measured by calculating the column efficiency (N)

expressed as the number of theoretical plates

in the column:

•

Columns that cause a lot of peak broadening have few theoretical plates.

•

Columns that produce very narrow peaks (little peak broadening) have a large number of theoretical plates.

Column Efficiency (N)

33

The three principle factors that cause band broadening and a decrease in column efficiency are described by the van Deemter equation:

e*

e

A ·

dp + B/µ

+ C ·

de2 ·

µH =Simplified version:

Particle size

Linear velocity (flow rate)Mobile phase viscosity

Particle sizeLinear velocity (flow rate)Mass Transfer

The Van Deemter Equation

34

Efficiency of a well-packed column will be a function of several factors, one of which will be particle size. As particle size decreases, efficiency will increase. In addition, back-pressure also increases as particle size decreases.

10 µm

50,000 P/m

5 µm

100,000 P/m

3 µm

150,000 P/m

sub-2 µm

300,000 P/m

Efficiency

Back-Pressure

Core-Shell

300,000 P/m

Effect of Particle Size on Efficiency

35

w1/2

w1/2

w1/2

5 µm

3 µm

Sub-2 µm

N = ~100,000 plates/mRelative BP = 1

N = ~150,000 plates/mRelative BP = 2x

N = ~300,000 plates/mRelative BP >4x

w1/2Core-ShellN = ~300,000 plates/mRelative BP ~3x

Effect of Particle Size on Efficiency

36

For a given particle size, column efficiency will be directly proportional to the length of the column. However, pressure

and elution time

will also be directly proportional to the column length.

25 cm 25,000 Plates40 min250 Bar

15 cm 15,000 Plates24 min150 Bar

10 cm 10,000 Plates16 min100 Bar

5 cm 5,000 Plates8 min50 Bar

Effect of Column Length on Efficiency

37

Column Length (mm)

Efficiency dp 5m

250 25,000

150 15,000

100 10,000

50 5,000

Balancing Column Length and Particle Size

Column Length (mm)

Efficiency dp 5m

Efficiency dp 3m

250 25,000 37,500

150 15,000 22,500

100 10,000 15,000

50 5,000 7,500

Column Length (mm)

Efficiency dp 5m

Efficiency dp 3m

Efficiency sub-2m / Core-shell

250 25,000 37,500

150 15,000 22,500 45,000

100 10,000 15,000 30,000

50 5,000 7,500 15,000

Column Length (mm)

Efficiency dp 5m

Efficiency dp 3m

Efficiency sub-2m / Core-shell

% Reduction in Analysis

Time

250 25,000 37,500

150 15,000 22,500 45,000 33

100 10,000 15,000 30,000 60

50 5,000 7,500 15,000 80

*Use shorter columns packed with smaller particles to reduce analysis time while maintaining/improving efficiency!!

38

Efficiency is also a function of mobile phase flow rate. All media will have an optimal flow rate, which is a function of particle size –

smaller particles have higher optimal flow rate.

A

B

Effect of Flow Rate on Efficiency

39

w1/2

tR

w1/2

tR

Very low flow

At or above optimum flow

Zero flow

Effect of Flow Rate on Efficiency

40

Effect of Flow Rate on Column Efficiency (100x4.6mm)

0

5000

10000

15000

20000

25000

30000

35000

0 0.5 1 1.5 2 2.5 3 3.5

Flow rate (ml/min)

N (P

late

s/co

lum

n)

Core-Shell 2.6

Luna 3u

Luna 5u

5µ

~1 mL/min

3µ

~1.5 mL/min

Core-Shell ~2 mL/min

Effect of Flow Rate on Efficiency

Core-Shell 2.6µ

Fully porous 3µ

Fully porous 3µ

41

The chromatographic measurements that we have discussed so far will all play a significant role in method development.

1. Capacity factor (k’)

–

retention of analyte relative to void t0

•

Controlled by modulating %strong mobile phase

2. Peak asymmetry

–

peak shape (fronting, tailing, symmetrical)•

Result of secondary interactions (e.g. Ionic in RP mode)•

Sensitive to sample loading & sample solvent effects

3. Selectivity (α)

–

difference in the k’

of two analytes•

Will be determined by mobile phase composition and nature of stationary phase

4. Efficiency (N)

–

function of peak width and retention•

Determined by particle size, column length, flow rate•

Column packing will affect efficiency (vendor)

Review

42

Resolution: The Goal of Chromatography

43

The goal and the purpose of liquid chromatography is to resolve the individual components

of a sample from each other so that they may be identified and/or quantitated.

Resolution: The Goal of Chromatography

44

Resolution is a measure of how well two peaks are separated from each other.It is calculated as the difference in retention time of two eluting peaks divided by the average width of the two peaks at the baseline.

R (resolution)

= RTB

- RTA

/ .5 (widthA

+ widthB

)

Resolution: The Goal of Chromatography

45

(1)

Retention time for the two peaks will be a function of capacity factor (k’).

(2)

The selectivity (α)

will also affect the retention time values for the two peaks.

(3)

Peak width will be a function of column efficiency (N)

and asymmetry (Asym).

k’

α

N, Asym

Resolution: The Goal of Chromatography

46

Selectivity

Capacity factorEfficiency

The equation below allows us to calculate the relative importance of each of these three factors in overall resolution:

It is important to note that you (the analyst) have control over

each of those factors through your choice of HPLC column and running conditions:

1.

Efficiency (N)

→

Particle size/morphology and column length2.

Selectivity (α)

→ Stationary phase and mobile phase3.

Capacity factor (k’)

→ Stationary phase and mobile phase

Resolution: The Goal of Chromatography

47

Ineffective after k’

~10

Constant increase in resolution

Most important determinant of resolution!!

The Effect of k’, α

and N on Resolution

48

Modulating column efficiency is a very effective way to optimize

resolution. There is a strong, linear correlation between N and Rs, but it is not a 1:1 ratio. Column efficiency is a flexible tool because we can easily modify it via changes in particle size

and column length.

Column Length (mm)

Efficiency 5m

Efficiency 3m

Efficiency sub-2m / Core-shell

250 25,000 37,500

150 15,000 22,500 45,000

100 10,000 15,000 30,000

50 5,000 7,500 15,000

More efficiency/resolution

Long

er R

unTi

me

More Back-Pressure

Mor

e Pr

essu

reImpact of Efficiency on Resolution

49

0

0.5

1

1.5

2

2.5

0 10000 20000 30000 40000

Relat

ive Re

solut

ion

Column Efficiency (Plates)

Doubling column efficiency increases Rs by a factor of 1.4x

Impact of Efficiency on Resolution

50

1.

For method development, start with an intermediate column length, packed with the smallest particle size that system pressure limitations will allow.•

Conventional HPLC 3 µm 150x4.6mm or core-shell 100x4.6mm•

UPLC sub-2 µm or core-shell particle•

Work at optimal flow rate for that particle size

2.

Fine-tune

for maximum productivity:•

Excessive resolution shorter column, increase flow rate•

Insufficient resolution longer column; modify flow rate to compensate for pressure

Optimizing Efficiency for Maximum Resolution

51

Capacity factor is the most important, yet limited, factor

in determining resolution. It is crucial to have a reasonable k’

value because analytes must be retained in order to separate them. The drawback is that at high k’

values, passive diffusion causes extensive band broadening and loss of performance.

The Impact of Capacity Factor on Resolution

52

1.

Adjust k’

value to be between 2 and 10•

In RP, adjust % of organic (acetonitrile or methanol)•

Altering nature of stationary phase/media can modulate k’

as well•

E.g. C18 versus C8 will give shift in k’

values

2.

At k’

< 2, have sub-optimal resolution•

May also have interference from solvent, non-retained components

3.

At k’

values > 10, band broadening due to diffusion limits resolution

gain•

In RP, complex mixtures of polar and non-polar components will require gradient for optimal performance/run time balance

•

Polar stationary phases can the “total elution window”

of complex mixtures in isocratic mode

Optimizing Capacity Factor

53

Small changes in selectivity can have a dramatic effect on retention. This is one of the reason why the same stationary phases

from different manufacturers can sometimes give very different results, and also why changes to mobile phase

composition can alter the results so strongly.

The Impact of Selectivity on Resolution

54

Method Development Exercise 1:   Optimization to Reduce Analysis Time and 

Increase Productivity

55

Column: C8 250 x 4.6mm 5µm

Mobile phase:

70 / 30 0.1M Ammonium acetate / THF

Flow rate:

1.0 mL/min

Components:

1-6 = Impurities A -

G7. Mupirocin

Mupirocin

Mupirocin Impurity Profile

min0 2 4 6 8 10 12 14 16

mAU

0

25

50

75

100

125

150

175

DAD1 A, Sig=240,10 Ref=off (Z:\1\DATA\MT050211\DJGSK 2011-05-02 15-34-44\MUPIROCIN000001.D)

1

2 3

5

6

7

~16min

Column:

5 µm

C8(2) 250 x 4.6mm

Mobile phase: 70/30

0.1M Ammonium acetate pH 5.7/THF

Flow rate:

1.0 mL/min

4

Rs 3/4 = 0.63; k’

= 2.3

56

Mupirocin: Original Method

57

Step 1. Reduce %organic to increase k’:•

Increases Rs•

Increases run time

Step 1. Adjust k’

to Increase Resolution

58

Column Length (mm)

Efficiency dp 5 m

Efficiency dp 3 m

Efficiency sub-2 m / Core-shell

% Reduction in Analysis

Time

250 25,000 37,500

150 15,000 22,500 45,000 33

100 10,000 15,000 30,000 60

50 5,000 7,500 15,000 80

Step 2. Switch to 150x4.6mm 3 µm media:•

Reduces analysis time•

Maintains efficiency

Step 2. Optimize Particle Size and Length

0

5000

10000

15000

20000

25000

30000

35000

0 0.5 1 1.5 2 2.5 3 3.5

Flow rate (ml/min)

N (P

late

s/co

lum

n)

Core-Shell 2.6

Luna 3u

Luna 5u

59

Step 3. Increase flow rate to 1.5 mL/min:•

Optimizes efficiency for 3 µm

Step 3. Optimize Flow Rate

min0 2 4 6 8 10 12 14 16 18

mAU

0

20

40

60

80

VWD1 A, Wavelength=240 nm (JL050211\MUPI0003.D)

1

23

4

5

6

7

Column:

3 µm

C8(2) 150 x 4.6mmMobile phase: 80/20

0.1M Ammonium acetate pH 5.7/THFFlow rate:

1.5 mL/min

Rs increased from 0.63 to 1.6Run time increased from 16 to 20 minutes

20 min

k’

= 9; Rs 3/4 = 1.6

60

Mupirocin: Intermediate Method

61

Column Length (mm)

Efficiency dp 5 m

Efficiency dp 3 m

Efficiency sub-2 m / Core-shell

% Reduction in Analysis

Time

250 25,000 37,500

150 15,000 22,500 45,000 33

100 10,000 15,000 30,000 60

50 5,000 7,500 15,000 80

Step 4. Switch to 100x4.6mm Core-Shell media:•

Reduce analysis time•

Increase efficiency

Step 4. Switch to Core‐Shell Media

min0 1 2 3 4 5 6 7 8 9

mAU

0

50

100

150

200

250

VWD1 A, Wavelength=240 nm (JL050211\MUPI0006.D)

Column:

Core-Shell 2.6 µm

C8 100 x 4.6mmMobile phase: 80/20

0.1M Ammonium acetate pH 5.7:THFFlow rate:

1.5 mL/min (2ml/min if pressure allows)

Rs increased from 1.6 to 2.3Run time decreased from 20 to 8 minutes

8 min

1

5

6

7

23

4

Rs

3/4 = 2.3

62

Mupirocin: Final Optimized Method

min0 1 2 3 4 5 6 7 8 9

mAU

0

50

100

150

200

250

VWD1 A, Wavelength=240 nm (JL050211\MUPI0006.D)

8 min1

5

6

7

23

4

Rs

3/4 = 2.3

Final Optimized Method:min0 2 4 6 8 1 0 12 14 16

m AU

0

25

50

75

1 00

1 25

1 50

1 75

D AD 1 A , S ig= 240,1 0 R ef=o ff ( Z:\1 \DAT A\MT 050 211\ DJ G SK 201 1-0 5-0 2 15 -34 -44 \MU PIRO C IN00 0001 .D)

12 3

5

6

7

16 min

4

Rs

3/4 = 0.63

Initial Method:

63

Mupirocin: Final Optimized Method

64

End of Part I