HPLC Method Development Flow Chart

HPLC Method Development Flow Chart

Handy Equations

start

backpressure

retention factor (k)

Legend

resolution (Rs)

= yes, proceed to next step

= an adjustment or decision to be made

Be aware of the physicochemical properties of the analytes

= no, change a condition

= something you need to pay attention to

selectivity (a)

theoretical plate height

= directional arrow; shows you next step of the method or directs you to a required decision

-this arrow can be followed when there is an alternative to 'yes' or 'no'

= criteria required to advance to next step

Select a column

= explanation / elaboration

Select the type of run

Isocratic

Gradient

make sure backpressure does not exceed 4000

This will damage the machine.

Start with a low flow rate and gradually increase.

Flow rate and back

pressure depend on

the column (and particle) dimensions.

implement 100% organic solvent

implement scouting gradient

A scouting gradient is a run with a gradient time of ~20mins over the organic range 5-100%

They help to determine if an isocratic run is possible, and if so at what %B

is K< 3?

compound is not suitable for analysis by reverse-phase HPLC

calculate the % B at half delta TG and conduct an isocratic run at this %B

adjust organic range to begin at the % B of first eluted peak

reduce % B by 10% increments

is resolution > 1.5?

Every 10% decrease in %B equates to an approximate 2-3 fold increase in K. Using this knowledge, a %B can be estimated with reasonable accuracy to increase K within method limits (<15)

adjust the gradient time Tg

Tg denotes the time taken to get from 5%B (0 minutes) to 100% (20 minutes)

change solvent but keep solvent strength constant with the use of a nomogram

The nomogram depicts conversions between solvent types whilst retaining the same solvent strength. For example, runs conducted with 50% ACN and 60% ACN should have approximately similar overall run times, whilst the selectivity of the two runs is likely to change.

This is due to different compounds engaging in different types and extents of interactions with different solvents based on their chemical properties.

For example, using a more acidic solvent is likely to decrease the retention time of a basic analyte due to ionisation.

adjust the organic range

is K less than 15 and resolution > 1.5?

This does not have to be done in a linear fashion. This method is very much trial and error.

is there observable/appreciable separation of all peaks after using two different mobile phases?

almost

is resolution > 1.5?

with the aid of a nomogram, choose an isoleluotropic solvent mixture that yields close to the target K

reduce the length of the column

Altering the length of the column changes linear velocity and plate count

Varying column length can help to optimise linear velocity and increases the plate count at the expense of increased backpressure and tR

is K less than 15 and resolution > 1.5?

is resolution > 1.5?

is resolution almost at target (>1.4) and run time acceptable?

almost

Increase temperature.

change organic solvent

Start editing efficiency.

Increasing temperature alters the pKa of the analytes, and to different extents.

The degree of ionisation is dependent on the chemical properties of the compound

Doing so decreases backpressure through viscosity reduction. This will allow for a higher flow rate implementation reducing the run time and potentially optimising Vm

changing the organic modifier allows for new interactions (and extents of) to occur between the analytes and the mobile phase; resulting in potentially major changes in selectivity

Efficiency method modifications are implemented when resolution is very close to acceptable but still not quite there (within 0.1 of Rs = 1.5). No major changes to resolution occur here.

almost

Change flow rate. Start at the maximum flow rate, then lower in increments.

Altering flow rate changes linear velocity.

By optimisation of FR the theoretical plate height can be reduced, increasing the plate count and henceforth efficiency.

is resolution > 1.5?

is K less than 15 and resolution > 1.5?

Adjust pH

is K less than 15 and resolution > 1.5?

Through this parameter the ionisation states of the compounds can be controlled and varied.

pH can only be between 2.5 and 8. Values outside of this destroy the beads/ligand

very close;

within 0.1 of Rs = 1.5

is resolution 1.5 or greater?

run time acceptable?

reduce the sample injection volume

Increase column length in increments.

this technique aims to limit longitudinal diffusion, and consequently reduce band broadening

Doing so alters Vm and increases the plate count.

Care must be taken to adjust flow rate with length increase, for an increase in column length corresponds to an increase in backpressure.

re-develop the method:

-change the column

-change to isocratic conditions

-change the solvent

is resolution > 1.5?

is resolution 1.5 or greater?

run time acceptable?

re-develop the method:

-change the solvent type

-change the column

if neither of these things work,

-try a gradient method

Decrease particle size

decreasing particle size reduces eddy diffusion and band broadening by limiting the variety of paths a compound can take within the column.

doing this also increases the theoretical plate height.

is resolution 1.5 or greater?

run time acceptable?

Decrease injection volume

this technique aims to limit longitudinal diffusion, and consequently reduce band broadening

is resolution > 1.5?

method validation

Limit of Detection

Range

Limit of Quantitation

Precision Testing

System Suitability

Quantification

Linearity

Accuracy Tests

Specificity

Robustness

Used to verify that the chromatographic system is adequate for the intended analysis

Being able to prove that any peak is caused by a lone compound without interference from degradation products or impurities.

For compound quantification, an external standard of (at least) 3 known concentration are injected and a calibration curve is derived from the AUC of each injection

Defined by concordance of test results when the method is applied to multiple samplings of a heterogenous sample.

Measured by RSD (Relative Standard Deviation)

This is the interval between the lowest and highest concentrations of analyte sample that were used in the linearity experiments to confirm that the method has acceptable levels of linearity, accuracy and precision.

Acceptance criteria: R squared > 0.99

50% to 150% of test concentration.

A measure of closeness of results yielded by a specific method to a known true value.

This is assessed using 3 concentration levels (minimum) (e.g. 50%, 100% and 150%), each in triplicate.

Reported as 'percentage recovery of known amount added'

Acceptance criteria: average recover 98-102%

The lowest concentration of analyte within a sample that can be detected but not necessarily quantitated

Acceptance criteria: Signal:Noise = 3:1

The lowest concentration of analyte within a sample that can be quantified with suitable accuracy and precision

Acceptance criteria: Signal:Noise = 10:1

A measure of the capacity of the analytical procedure to remain unaffected by small, deliberate variations in method parameters. This provides an indication of reliability during normal usage.

Plackett Burman acceptance criteria:

Meet system suitability and %RSD requirements for all experiments and samples (<20% difference from the original method conditions)

Can be confirmed by analysing varying concentrations of a compound (over 5 concentrations, 12.5% - 150%) and plotting the results.

Analysing different concentration should yield predictable results with an R squared value of 0.99 or greater.

Below this is unacceptable.

Acceptability Criteria:

Resolution between API and nearest impurity > 2.5

tR ~ the peak should be well-resolved from the void volume ( k > 2 )

Overall standard precision/repeatability (n>5) for AUC, tR and RSD (<1%)

Tailing factor < 2

Plate count (N) > 10,000

Calibration curve R squared of 0.995

This can be done by instigating degradation of analytes and spiking the sample with suspect impurities. Resolved identifiable peaks account for acceptable specificity

Reproducibility

Repeatability

Intermediate Precision

Precision under variations within the same laboratory (days, analysts, equipment)

This is the ability to reproduce data within the predefined precision across two (or more) different laboratories

Precision under the same operating conditions over a short interval of time

A minimum of 6 determinations at the 100% level

Acceptance criteria: RSD<2% on individual basis of 2% (RSD<2% overall)

Acceptance criteria: RSD<2%

resolution > 1.5?

run time < 15 minutes?

sensitivity > 1?

backpressure < 4000 psi?

(<9000 if using UPLC)

Successful method.