No thanks! I would like to know more about CHROMacademy

 Over 3000 E-Learning topics / 5000 Articles & Applications
 

The CHROMacademy Essential Guide Webcast:
Developing Better HPLC (MS) Methods

Thursday 26th September 2013, 16:00 GMT

Where do you start when you want to develop a better HPLC method? Whether you work in a regulated environment or not, setting specifications for your HPLC protocols is a good practice. In this Essential Guide we will set out guidelines for chromatographic parameters such as retention, resolution, and efficiency that can be used to improve your HPLC methods prior to method validation. We will explain how mobile phase design and instrument parameters can be used to achieve improvements in method robustness and what to look for when a method is failing. Practical tips for sample and eluent preparation, and the correct detector settings to use will also be discussed.

Kevin Schug (Assistant Professor, University of Texas, Arlington) and Dawn Watson (CHROMacademy Technical Expert, Crawford Scientific) will look at the major factors that affect chromatographic robustness to give you the tools to develop better methods that will give consistent, reproducible chromatography.

 

Topics covered include:

  • What target specifications should be set for an HPLC method
  • How sample preparation can impact on the sensitivity and robustness of an HPLC method
  • Stationary phase chemistry and important column properties
  • Designing ‘smart’ eluent solutions
  • Detector parameters to improve quality and robustness
  • Developing inherently robust methods
  • Aspects of LC-MS methods which can (and should!) be optimized

Who Should Attend:

  • Anyone working with HPLC in a regulated environment
  • Anyone who wants to develop better HPLC methods
  • Anyone who wants practical advice on improving their HPLC separations
  • Anyone who wants to better understand the factors that affect an HPLC separation
  • Anyone working with LC-MS
 


Find out more about this Month's Essential Guide Webcast »

The Camtasia Studio video content presented here requires JavaScript to be enabled and the latest version of the Adobe Flash Player. If you are using a browser with JavaScript disabled please enable it now. Otherwise, please update your version of the free Adobe Flash Player by downloading here.

The CHROMacademy Essential Guide Tutorial
Developing Better HPLC (MS) Methods

Where do you start when you want to develop a better HPLC method?  Whether you work in a regulated environment or not, setting specifications for your HPLC protocols is a good practice.  In this Essential Guide we will set out guidelines for chromatographic parameters such as retention, resolution, and efficiency that can be used to improve your HPLC methods prior to method validation.  We will explain how mobile phase design and instrument parameters can be used to achieve improvements in method robustness and what to look for when a method is failing.  Practical tips for sample and eluent preparation, and the correct detector settings to use will also be discussed.

Dr. Kevin Schug (Assistant Professor, University of Texas, Arlington) and Dr. Dawn Watson (CHROMacademy Technical Expert, Crawford Scientific) will look at the major factors that affect chromatographic robustness to give you the tools to develop better methods that will give consistent, reproducible chromatography.

Topics covered include:

  • What target specifications should be set for an HPLC method
  • How sample preparation can impact on the sensitivity and robustness of an HPLC method
  • Stationary phase chemistry and important column properties
  • Designing ‘smart’ eluent solutions
  • Detector parameters to improve quality and robustness
  • Developing inherently robust methods
  • Aspects of LC-MS methods which can (and should!) be optimized

 

U.S. Food Safety Training

“Upping the Ante: Food Safety Training”

EU Food Safety Training.

“Fera IFSTL Overview”

Share this tutorial

Ultimately sensitive, reproducible, and robust chromatographic results, which are fit for purpose, according to the analytical requirements are desired.  This may mean different things to different people, depending on the type of work that is being carried out.  Having chromatographic performance targets to work towards will not only result in more robust chromatography but they will be a great indicator of when wrong development path has been chosen, or when there are underlying problems with the method or equipment.

There are many regulatory organisations that have set analytical method performance requirements.  Table 1 shows the current US Food and Drug Administration (FDA) values for the Validation of Chromatographic Methods which will be used for the purposes of this discussion.

 
 

 

Download - FDA Validation of Chromatographic Methods

Download - ICH Harmonised Tripartite Guideline Validation Of Analytical Procedures: Text And Methodology Q2(R1)

 
Table 1: The current FDA values for the Validation of Chromatographic Methods.
 

Retention Factor

2 < k < 10

It is suggested that values of k should be between 2 and 10, but this may not work in all cases.  If k < 1 separations will be less stable and reproducible.  They will have a greater susceptibility to chromatographic interferences at the beginning of the chromatogram, with the chance of peaks being poorly resolved from unretained material at t0.  The retention of analytes with low k values will also be more sensitive to small changes in mobile phase composition.  However, sometimes a k value of between 1 and 2 may work well when faster chromatography is desired (high throughput), where samples do not contain a lot of endogenous/matrix components or where pH or buffer strength are not vital in controlling retention or selectivity.

Caution
Always check specificity and reproducibility when working a 1 < k < 2

 

For complex mixtures k values greater than 10 may be needed to resolve all peaks.  If complex mixtures are being analyzed be aware of peak broadening of later eluting peaks that may reduce resolution.

The largest gain in resolution is achieved when the k value is between 1 and 5 (Figure 1).  Above a k value of approximately 5 increasing retention only provides minimal increases in resolution.  Too much retention wastes valuable analysis time and the chromatographic peak height will decrease as the bandwidth of the peak increases.  This is due to HPLC being a diffusion limited technique and above k values of approximately 10 increasing retention will have little or no effect on resolution because of the decrease in efficiency resulting from an increase in longitudinal molecular diffusion.     

 

Figure 1: Plot of k vs. resolution.

 
 

Longitudinal Diffusion

A band of analyte molecules contained in the injection solvent will tend to disperse in every direction due to the concentration gradient at the outer edges of the band.  This broadening effect is called longitudinal diffusion because inside tubes the greatest scope for broadening is along the axis of flow (Figure 2).  The band will broaden in all system tubing, but the worst effects will be encountered in the column itself.  Longitudinal diffusion occurs whenever the HPLC system contains internal volumes that are larger than necessary.  Longitudinal diffusion has a much larger effect at low mobile phase velocity (flow), therefore, using high linear velocity (high mobile phase flow with narrow columns) will reduce the effect of this broadening.

 

Content on this page requires a newer version of Adobe Flash Player.

Get Adobe Flash player

Figure 2: Longitudinal diffusion.
 
 

Real world example

An example in modern chromatography which demonstrates that sometimes larger k values are acceptable, and in this case necessary, in order to generate the required resolution is the separation of 8 nitro aromatic compounds (Figure 3 – 5).  At 80 %B (Figure 3) the last eluting peak has a k of 1.09 and only 3 analytes are fully resolved, reducing the %B to 70% increases the k value to 5, however, there are still only 7 of the 8 compounds resolved (Figure 4).  A k of 10 (60% B, Figure 5) is required to fully resolve all 8 compounds. 

The most effective and convenient way to alter the retention factor of a peak is to adjust the solvent strength of the mobile phase.  This is usually achieved in reversed phase chromatography by changing the amount of organic solvent in the mobile phase.  For reversed phase HPLC the stationary phase is non-polar, therefore, increasing the polarity of the mobile phase will increasingly repel the hydrophobic (non-polar) sections of the analyte into the stationary phase and the analyte will be retained for longer on the column.  The converse is also true.  An increase in the organic content of the mobile phase of 10% decreases the k for each component by a factor of 2 to 3.

According to the fundamental resolution equation (Equation 2) resolution increases when retention (k) increases.  The nitro aromatic analytes in the example below are relatively polar, therefore, a highly polar mobile phase will decrease the retention of these analytes, hence, decreasing resolution (i.e. 80% MeCN Figure 3).  Therefore, a weaker mobile phase is required to obtain the resolution required to separate the analytes.

 

 

Retention Factor

Figure 3: Separation of 8 nitro aromatic compounds.  Column: C18 10 cm x 0.21 cm x 2 μm.  Eluent: Water:MeCN (80% B).  Flow rate: 1.0 mL/min. 
Temperature: 35 oC.  For the final peak (highlighted) k = 1.09, N = 21629, and Rs (min) 0.05. 

 
 
Retention Factor

Figure 4: Separation of 8 nitro aromatic compounds.  Column: C18 10 cm x 0.21 cm x 2 μm.  Eluent: Water:MeCN (70% B).  Flow rate: 1.0 mL/min. 
Temperature: 35 oC.  For the final peak (highlighted) k = 5, N = 21182, and Rs (min) 0.29. 

 
 
Retention Factor

Figure 5: Separation of 8 nitro aromatic compounds.  Column: C18 10 cm x 0.21 cm x 2 μm.  Eluent: Water:MeCN (60% B).  Flow rate: 1.0 mL/min. 
Temperature: 35 oC.  For the final peak (highlighted) k = 10, N = 21030, and Rs (min) 2.03. 

 
 

Efficiency

Where:

tR = retention time
Wb = peak width
W1/2 = peak width at half height
L = column length
H = height of a theoretical plate

N > 2000

 

Efficiency can be increased by increasing the column length, reducing the column internal diameter, or decreasing the particle size.  It is better to use a smaller diameter packing than increase the column length, which will increase analysis time.  However, a decrease in particle size will result in an increase in system backpressure.  The use of smaller particles and narrower column internal diameter both require minimized extra column dead volume in order to avoid efficiency losses.

The FDA stipulates a value for N > 2000 which is typically easily achieved with modern HPLC columns (Table 2).

 
Table 2:  Approximate column efficiencies for standard HPLC column geometries.
 
 

A very practical measure of the relative separating power of columns of different dimension is the so-called Resolution Capacity (Table 3).  This can be calculated for a column by dividing the length by the particle size.  Table 3 shows some values of resolution capacity for some common column geometries.  When comparing, designing, or translating methods this can be a useful measure to determine whether the same, or increased, resolution can be achieved.  For example, to retain the amount of resolution from a 300 mm column with 10 μm particles (resolution capacity 30,000) a shorter 150 mm column with 5 μm particles could be used (Figure 6).  Or to increase resolution a 150 mm column with 1.7 μm particles could be used instead of a 100 mm column with the same particle size, this would result in a n increase in resolution capacity from 58,820 to 88,230 respectively.

 

Table 3: Resolution capacity of standard HPLC column geometries.

 
 

Figure 6: Separation of a mixture of 6 test compounds over a wide range of LogP values.  Column:  Type II C8.  Mobile phase: Water/Acetonitrile.  Flow rate: 2 mL/min.

 
 

Tailing Factor

T ≤ 2

Tailing peaks create issues with resolution, quantitation (integration), and reproducibility (Figure 7).  Peak shape is often the controlling factor when optimising complex separations, especially when components are present in very different concentrations.

 

Figure 7: TF = (a) 1.24, (b) 1.42, (c) 1.58 and (d) overlay of a, b, and c. 1

 
 

Real world example

The analysis of a stability indicating sample at different buffer concentrations demonstrates the importance of the tailing factor (Figure 8 and 9). At a lower buffer concentration (13.4 mM) the degradent peak in the sample has a relatively high tailing factor of 1.72, which although it is in the FDA recommended range, results in poor resolution (Rs = 0.98) between the degradent and the proceeding peak. Increasing the buffer concentration (23.6 mM) not only improves the peak tailing of the degradent peak (TF = 1.43) but also results in resolution of the two peaks (Rs = 1.69). Note: The retention time of the peaks does not change. The improvement in peak shape is due to the increases in buffer concentration.

Reversed phase separation of ionisable analytes is particularly susceptible to changes in pH which is why buffers are used to carefully control and stabilise the pH of the mobile phase. There are cases where analyte retention in reversed phase HPLC is affected by buffer concentration. These cases are usually confined to situations where there are ion exchange interactions taking place between basic solutes and acidic silanols on the surface of the silica stationary phase. Above pH 3 silanol groups on the silica surface will be ionized and can potentially interact with analyte molecules resulting in changes in retention and peak shape. An increase in buffer concentration, and hence an increase in ionic strength, of the mobile phase can suppress this ion exchange interaction due to increasing competition from the buffer counter ions. It is also worthy to note that as buffer concentration is increased the mobile phase is made more polar (ionic) which can affect analytes in differing ways depending on the analyte chemistry; some analytes may experience reduced retention, some slightly more.

 

Figure 8: Stability indicating sample (no ionizable compounds), separation previously optimized for selectivity.  Column: C18 10 cm x 0.21 cm x 2 μm. 
Eluent: Phosphate buffer (aq):MeOH.  Flow rate: 1.0 mL/min.  Temperature: 35 oC.  Buffer concentration = 13.4 mM.  Rs = 0.98.  Degradent TF = 1.72.

 

Figure 9: Stability indicating sample (no ionizable compounds), separation previously optimized for selectivity.  Column: C18 10 cm x 0.21 cm x 2 μm. 
Eluent: Phosphate buffer (aq):MeOH.  Flow rate: 1.0 mL/min.  Temperature: 35 oC.  Buffer concentration = 23.6 mM.  Rs = 1.69.  Degradent TF = 1.43.

 
 

Precision

RSD of ≤ 1% for n ≥ 5
n = number of injections

Injection precision is important for reproducible chromatographic results and should be estimated in the same way for each analysis.  It is indicative of performance of the plumbing, column, and environmental conditions at the time of analysis and assessment of injection reproducibility can be used to aid in the diagnosis of potential system problems such as leaks (Table 4).  It is expressed as RSD (relative standard deviation) and is measured by multiple injections of a homogeneous sample (Table 5).  Modern autosamplers injection precision is in the range 0.15 – 0.9% (depending on the sample volume).

 

Table 4: Injection repeatability data for an HPLC system that developed a leak during sampling.

 

Table 5: Representative data for 10 injections.

 
 

There are several factors that will affect precision which should be considered to minimise errors.

pH/Buffer type – When the pH of the mobile phase is close to the pKa of the analyte molecules peak shape will be poor and small changes in pH (as little 0.1 unit) can have a major effect on the retention and resolution of the compounds.  Buffer type is an important consideration.  A particular buffer is only reliable at 1 pH unit either side of its pKa (Table 10) and volatile buffers should be used for LC-MS applications to avoid fouling of the API source.

O2 – Some analytes will be susceptible to oxidation by atmospheric oxygen resulting in degradation and poor reproducibility.  

Light – Light sensitive samples that degrade on standing will also produce erroneous chromatographic peaks that will result in poor precision and reproducibility.  Amber vials can be used for light sensitive samples; however, these vials will often have a high iron content which may lead to further sample degradation and repeatability issues.

Heat – To mitigate loss of heat sensitive samples thermostated autosamplers can be used to keep samples at lowered temperatures.

Organic components – HPLC grade solvents and reagents should always be used.  Water should be a free solvent; however, high purity water is required for all sample and mobile phase preparation protocols in HPLC.  Poor quality solvents, reagents, and water can produce a multitude of chromatographic errors including; altered resolution, ghost peaks, changes in stationary phase chemistry and baseline issues.  Possible sources of organic contaminants are from feed water (i.e. tap water), leached from purification media (syringe filters), tubing and containers, bacterial contamination, and potentially, absorption from the atmosphere. For example, alkaline mobile phases are known to absorb polar organics such as formaldehyde, amines, and atmospheric carbon dioxide.  Organic contaminants with UV active chromophores can interfere with quantification if they are present in high enough levels.  Reactive contaminants may also produce unwanted side reactions with analyte molecules.

Mobile phase stability – In the presence of air or oxidisers THF will also form hazardous, explosive peroxide species, which pose both a safety risk and can be reactive towards analytes.  Care should also be taken if reactive analytes, (alcohols, aldehydes, carboxylic acids) are being analyzed as in the presence of MeOH methyl esters can be formed giving rise to erroneous peaks in the chromatogram and quantification errors.  If volatile additives, such as TFA, are being used the mobile phase reagent bottles should be capped to avoid evaporation of these volatile additives which would result in a change in the mobile phase composition and, hence, the retention characteristics of the analytes.

Adsorption – Adsorption to sample, containers, syringes, sample loops etc. will result in injection of different concentrations for each analytical run which will impact reproducibility of a method.

Dilutions – Many samples will not be in the correct format for analysis and may require dilution so as to avoid column overload.  These dilutions need to be carried out accurately for each sample to avoid errors in quantitation and reproducibility.

Sample mixing – The injected sample should be representative of the sample that is being analyzed, therefore, samples should be thoroughly mixed.

Homogeneous sampling – As in the example above each injection should be representative of the sample being analyzed.

Cavitation – Overfilled vials can lead to cavitation.  Cavitation is the formation of cavities in a liquid (i.e. bubbles or voids).  These are the consequence of forces acting upon the liquid and usually occur when a liquid is subjected to rapid changes in pressure.  In the case of over filled HPLC vials there is no headspace and a vacuum can be created when the sample is aspirated by the autosampler which will result in an inaccurate volume of sample being drawn up, again causing issues with reproducibility.

Emulsions – If a sample is an emulsion it will settle upon standing, consequently, a representative sample will not be injected onto the HPLC column and errors in quantitation will occur between samples.  Some liquid handling platforms may have the option of a stirrer block or shaker which could be employed to negate the problem of settling emulsions.

Internal standards – Internal standards can be used to estimate sample loss during sample pre-treatment.  The internal standard chosen should have similar behaviour to the analyte of interest under the pre-treatment steps.  An internal standard can also compensate for changes in sample size and concentration due to variations between instruments.

Most of the points above relate to injection precision.  Precision of a method can be also be affected by sample preparation techniques, instrument reproducibility (including injection techniques), S/N for the peak of interest, data handling, and the method of quantitation or calibration.

 
 

Resolution

Rs > 2

Before setting a value for resolution it is important to ask a couple of questions that relate to the specific separation first:

  • What value is acceptable?
  • What value is required for reliable quantitation?

Resolution is a function of retention (k), selectivity (α), and efficiency (N) which can all be modified, as has been seen, to improve resolution.  Rs > 1.5 can often be easily obtained for samples containing 5 or less components, however, for complex mixtures Rs > 1.8 is required for rugged performance.  ‘Real world’ setting of resolution specifications requires experience in HPLC and the method under consideration.  It can be useful to ask some questions before setting parameters for a separation.

  • Is the function of the analysis to be separation subject to qualitative or quantitative treatment, or is it to recover pure sample fractions
  • Is information available on the chemical identity of each of the sample components or will some qualitative analysis be necessary
  • Do all sample components need to be resolved
    • In some situations it is not necessary to separate every component and development may be concentrated in the separation of one or two of the components from the rest of the sample matrix components i.e. separation of drug and one impurity from the soluble excipients within a tablet mixture
  • What precision and accuracy is required for quantitative analysis
    • Typically between 0.5 and 2.0% precision can be achieved
  • How many different sample matrices will the method need to be developed for i.e. active ingredients in tablets, sterile ampoules, creams etc. Do the different matrices dictate that separate HPLC methods are used / is a single procedure more desirable.
  • How many samples will be analyzed in any one campaign
    • This will dictate the range of k' values which are appropriate for the analysis i.e. resolution or efficiency may be traded for shorter run times where short run times are required
  • What HPLC equipment will be available for the assay when it is routine
    • Is temperature control available, will the instrument be capable of gradient formation, what extra column effects will there be?
 
 

Real world example 1

Although values of Rs > 2 are recommended this is sometimes not practical with very complex samples, or depending on the type or stage of analysis.  The early stage analysis of pharmaceutical impurities (Figure 10) was carried out with a minimum Rs value of 1.2 which, for this type of analysis, gave reliable, reproducible quantitation of all impurities.  In the case it was required that the method was developed quickly.  A quick reporting limit of 0.05% or above was assigned to the method and only once this limit was reached was any further method development required.  This analysis was fit for its purpose and gave quick chromatographic results. 

 

Figure 10: Early phase pharmaceutical impurity analyses.

 
 

Real world example 2

Conversely, late stage pharmaceutical impurity analyses (Figure 11) required much more stringent Rs values (> 4) to give reliable quantitation of impurities on the tail of the API.  This kind of method will be used for regulatory filing so stringent values of Rs are required.  Any integration differences will result in issues with method robustness.  Due to the tailing API peak resolution and integration of the two small impurity peaks, which are also close together, is problematic.  The tailing peak of the API will give false values for Rs for both impurity peaks. 

 

Figure 11: Late phase pharmaceutical impurity analyses.

 
 

Reliable determination of the resolution factor Rs requires that the two peaks are resolved from each other, hence, in the case outlined above a reliable value for resolution cannot be obtained as the baseline widths will be affected by the amount of interference of the larger peaks with the smaller ones.  Another disadvantage of the resolution value is that it does not take into account the relative peak heights (Eqn. 6).  This can result in Rs values reaching values in excess of 1.5, which is considered as well resolved peaks, however, only one peak may be observed if the peak heights are significantly different.  In this case the use of the resolution equation will not give satisfactory results and other more descriptive measures such as the discrimination factor should be employed and can be used even when the maximum of the major peak is out of scale which may occur in trace analysis.

 

Discrimination factor

With complex samples it may be worth considering the use of a more descriptive measure such as the Discrimination Factor (d0, Equation 7, Figure 12).  The discrimination factor is often used when the peak heights are significantly different.  It is the ratio of the height of the valley to the height of the smaller peak (Equation 6).  If the value of d0 is 1 the peaks are completely resolved, however, when the value equals 0 the small peak has disappeared into the larger peak. 15

Where:

hv = height of the valley
hp = height of the smaller peak

 
Figure 12: Schematic representation of discrimination factor d0.
 
 

Analytical Chemists

  • I feel empowered to fix things
  • I can troubleshoot effectively
  • I know where to go for help
  • I understand my analyses
  • I know where to get applications
  • I’m up to date
  • I’m more employable
  • My career is progressing
 

Laboratory Managers

  • Improved equipment utilization
  • Faster method development/problem solving
  • Flexible workforce with a common standard
  • Better quality data
  • Get up to speed quicker
  • Lower T&E
  • Less reliant on me
  • I spend less time on training
 

Subscribe for $399 per/year and access:

  • The entire e-Learning archive
  • All Essential Guide Webcasts and Tutorials
  • LCGCs archive of articles and webcasts
  • Expert troubleshooting advice when needed
subscribe now
 
 


Like CHROMacademy on Facebook and keep up-to-date with the latest webcast, tutorial and eLearning release schedules.

 

Facebook

Sample preparation is often required in HPLC to obtain the sample in a matrix that is as close to the elution solvent as possible.  The types of sample preparation that are commonly used are:

  • Direct injection (i.e. no sample preparation)
  • Filtration or centrifugation
  • Dilution and injection
  • Solid phase extraction (SPE)
  • Liquid-liquid extraction
  • Soxhlet extraction
  • Solvent exchange

Methods 1 – 3 are often preferable as the sample is only slightly altered, they can be performed in a short preparation time, and the risk of analyte loss is reduced.  Methods 4 to 7 often involve multistep protocols which increases the chance of error.  There is further information on many of these methods in CHROMacademy so they will not be discussed fully here.  It is of course prudent to note that any sample preparation method should be optimized for each application and should remain consistent for subsequent analyses to ensure robust, reproducible results.

There are myriad advantages to optimising and performing sample preparation including, but not limited to:

  • Helps purify the sample
  • Increases analyte concentration
  • Reduces matrix interference problems
  • Improves chromatographic behaviour of the analyte
  • Provides robust, reproducible methods
  • Removes unwanted matrix material
  • Eliminates compounds that can cause ion suppression in LC-MS (TFA, phospholipids)
  • Can protect HPLC system and column from contamination and damage
 
 

Mass Loading Considerations

The baseline width of a chromatographic peak (W) determines its separation from adjacent peaks.  For an overloaded separation this can be related to the sample weight and experimental conditions for isocratic elution (Equation 8). 

Note: The first term of equation 8 relates to column effects while the second term relates to sample weight effects.  Therefore, for small samples peak width is determined by k, N, and t0, whereas, for large sample weights it is determined mainly by k, t0, and sample weight.  Rearrangement of Equation 8 to solve for the maximum column capacity (Ws) and plotting of these terms gives a graph with a slope with the value of Ws (Figure 13).  For different amounts of loaded sample, different values of (k0), (N0) and (W) will be found.

Where:   

Ws = column capacity
Wx = weight of compound injected
W = peak width at base
t0 = column void volume
k0 = retention factor

 

Figure 13: Plot of the two terms from Equation 8.  Slope of the line gives a value for Ws.

 

Even if the sample injection volume is small mass overload of the column can still occur resulting in fronting, tailing, or broadened peaks, a change in the retention time, or efficiency (Figure 14).  This is due to columns having a limited capacity (Table 6).  Mass overload will result in the stationary phase becoming saturated with sample.  Mass overload is likely to occur if the concentration of an analyte changes from sample to sample.  The effect of changes in analyte concentration should be evaluated for an HPLC procedure after method development and a maximum analyte concentration should be established, Wmax (Equation 9).  If a sample exceeds this limit it should be diluted and re-run.  Equation 9 can be used to estimate values of Wmax, which will be in micrograms (μg) and is independent of column length.  The value of Wmax is for each analyte in the sample, not the total sample weight; therefore, this value can be multiplied by the number of analytes in the sample to give the maximum sample weight.

Where:

k = retention factor
dc = column diameter (cm)

 
 

Worked example

For a separation performed on a 4.6 mm diameter column with an analyte k = 5.

 

Table 6: HPLC column mass capacities.  μg Sample on column.  Divide by the number of components for ‘per analyte loading’.

 
 

Figure 14:  Peak fronting caused by column overload.

 
 

Sample Solvent and Volume

Ideally samples should be dissolved in the same solvent as the mobile phase.  However, realistically this may not always be possible due to solubility issues.  Insolubility will result in peak tailing and ultimately blockage of the HPLC column.  Use of a weaker solvent with a similar chemical composition can be advantageous as it may focus the analyte at the head of the column giving greater peak efficiency.  Example of suitable weaker solvents are pentane for elution with hexane (normal phase) or water for elution with mixtures of MeOH water (reversed phase).  If the sample solvent is stronger peak broadening and poor peak shape (split, fronting, tailing) may be observed (Figure 15).  The volume injected should be kept as small as possible to minimise band broadening effects.  Maximum injection volumes depending on the sample solvent strength are detailed in table 7.

 

Table 7: Maximum sample injection volumes depending on solvent strength.

 
 
Figure 15: Sample dissolved in (A) 33% acetonitrile (B) 66% acetonitrile and (C) 100% acetonitrile. Caffeine at 0.75 mg/mL, 4 μl injection volume.  Column: Luna 3 μm C18(2) 50 × 2.0 mm. Mobile phase: gradient 5 to 95% acetonitrile (with 0.1% formic acid) in 7.5 min at a flow-rate of 1 mL/min.
 
 

Analytical Chemists

  • I feel empowered to fix things
  • I can troubleshoot effectively
  • I know where to go for help
  • I understand my analyses
  • I know where to get applications
  • I’m up to date
  • I’m more employable
  • My career is progressing
 

Laboratory Managers

  • Improved equipment utilization
  • Faster method development/problem solving
  • Flexible workforce with a common standard
  • Better quality data
  • Get up to speed quicker
  • Lower T&E
  • Less reliant on me
  • I spend less time on training
 

Subscribe for $399 per/year and access:

  • The entire e-Learning archive
  • All Essential Guide Webcasts and Tutorials
  • LCGCs archive of articles and webcasts
  • Expert troubleshooting advice when needed
subscribe now
 
 


Like CHROMacademy on Facebook and keep up-to-date with the latest webcast, tutorial and eLearning release schedules.

 

Facebook

Solvent Selection

Solvent selection may be one of the most important parameters in an HPLC separation due to the effect it can have on the selectivity.  In fact selectivity may be the most effective tool for optimising resolution (Figure 16).  Each solvent will interact differently with differing analytes and can be classified by their solvochromatic parameters (Table 8).  Dipole character π*, is a measure of the ability of the solvent to interact with a solute via dipolar and polarisation forces and will be good for the elution of polarisable analytes.  Acidity α, is a measure of the ability of the solvent to act as a hydrogen bond donor towards basic (acceptor) solutes so will be good for the elution of bases.   Basicity β, is a measure of the ability of the solvent to act as a hydrogen bond acceptor towards an acidic (donor solute), therefore, it will elute acidic analytes well.  These characteristics, along with knowledge of the analyte chemistry, can be used to manipulate elution. 

When deciding on the organic modifier that will be used for a separation the following points should also be considered.  The chosen organic modifier must be miscible with water (MeOH, THF, MeCN are all water miscible).  The use of a low viscosity solvent is preferable (Table 9) due to the lower pressure drop produced at a specific flow rate.  It also allows for faster chromatography due to the increased rate of mass transfer.  It should be noted that the viscosity of mobile phase mixtures will be markedly higher, and also therefore instrument pressure, than the pure compounds.  Figure 17 illustrates the situation with different percentages of methanol, acetonitrile and tetrahydrofuran in aqueous.  The viscosity maximum for MeOH/Aq. mixtures is reached at 40% MeOH (1.62 mPa s at 25 oC) which is almost three times the value for MeOH alone.  Acetonitrile is often the solvent of choice due to its ability to solubilise many small molecules and its low viscosity.  As can be seen from Figure 17, the viscosity of MeCN/Aq. mixtures decreases with increasing amounts of MeCN, with the maximum value being ~0.95 mPa S (20 %B at 25 oC) which is almost half the maximum of MeOH/Aq.  Increased solvent viscosity will result in higher system pressures.  It is not uncommon for an instrument to shut down due to over pressuring when employing methanol as the organic solvent during the first injection of gradient method.

When mixing MeOH and aqueous mixtures, each solvent should be weighed or volumetrically measured due to the solvent contraction that occurs upon mixing i.e. 500 mL of water topped up to 1000 mL with MeOH will result in a solution with a MeOH content in excess of 50% by volume.  Care should also be taken if reactive analytes, (alcohols, aldehydes, carboxylic acids) are being analyzed as in the presence of MeOH methyl esters can be formed giving rise to erroneous peaks in the chromatogram and quantification errors.

Tetrahydrofuran is an interesting solvent in that it is one of the strongest chromatographically and can produce separations in very short times, whilst still being fully miscible with water.  However, it does have a relatively high UV cut off.  Column equilibration can also be slower with THF than with MeOH or MeCN.  In the presence of air or oxidisers THF will also form hazardous, explosive peroxide species, which pose both a safety risk and can be reactive towards analytes.  Recently THF has also been recently upgraded to carcinogen status by some bodies.  

If a UV detector is being used it is important to consider the UV cut off of the mobile phase (organic modifier, buffers, additives etc.) to ensure that they do not interfere with the λmax of the analyte.

 

Figure 16: The impact of selectivity on resolution.

 
 

Table 8: Solvochromatic parameters of common HPLC solvents. The values for π*, α, and β are normalized so that their sum gives 1 and are therefore only relative numbers.

 
 

Table 9: Properties of common reversed phase HPLC solvents.

 
 

Figure 17: Viscosity of mixtures of aqueous and organic solvents at 25 oC. 14

 
 

LC-MS Applications

LC-MS applications require special consideration to optimize the mobile phase and achieve sensitive MS detection of analytes.

Electrospray Ionisation (ESI)
The solvent should support ions in solution, i.e. a solvent with some dipole moment.  Solvents that are more viscous are less volatile and will reduce sensitivity.  A higher percentage of organic modifier gives better sensitivity due to the decreased surface tension and lower solvation energies for polar analytes.  Reversed phase solvents are suitable as they are often polar, whereas, normal phase solvents are not.

Atmospheric Pressure Chemical Ionisation (APCI)
Most solvents are compatible.  It is preferable to have neutral molecules.

Other components of the mobile phase should also be considered.  For example non-volatile additives will crystallise and block the ion source so these are best avoided e.g. surfactants such as triton and sodium dodecyl sulphate.  Compounds that reduce ionization/ion formation (DMSO, TRIS, glycerol) are also not recommended for use with LC-MS applications.  Corrosive reagents such as inorganic acids (H2SO4, H3PO4) or alkali metal bases (NaOH, KOH) will damage equipment.  Strong ion pair reagents will cause ion suppression (TFA).  However, additives such as triethyl- and trimethylamine may enhance ion formation in negative mode.

 
 

Ionisable Analytes: Extent of Ionisation

The change in degree of ionization of an ionisable molecule happens over a limited pH range.  Due to pH and pKa being logarithmic values it can be shown that 1 pH unit away from the pKa the extent of ionization is approximately 90%.  At 2 pH units away from the pKa the extent of ionization is approximately 99%, while at 3 pH units it is 99.9%.  A good rule of thumb to predict the extent of analyte ionization is the 2 pH rule (Figure 18).

 

Figure 18: 2 pH rule for determining the extent of ionization of acidic and basic analytes.

 
 

As can be seen from Figure 19 and 20 a small change in pH can have a large impact on the resolution of a separation.  Figure 19 shows the analysis of a mixture of acids and neutrals at pH 4.1.  At this pH the acidic analytes are ionized and are, therefore, polar which results in poor retention and resolution  A reduction in pH to 3.0 (Figure 20) leads to the acidic analytes being protonated to give the neutral species (ion suppressed) which are less polar than the ionized form, therefore, improving both the retention and resolution of theses analytes.

 

Figure 19: Analysis of a mixture of acids and neutrals at pH 4.1.  Column: C18 100 mm x 2.1 mm x 1.7 μm.

 
 

Figure 20: Analysis of a mixture of acids and neutrals at pH 3.0.  Column: C18 100 mm x 2.1 mm x 1.7 μm.

 

Basic analytes often produce chromatograms with poor peak shapes.  This is due to the interaction with surface silanol species (Figure 21).  Basic analytes cannot be analyzed using ion suppression techniques as the pH increase required to render these species neutral, and hence disrupt the surface interaction, may damage the silica column. 
There are several approaches to improving the peak shape of basic analytes, including, the use of sacrificial bases, modified silica, or low pH (< 3).  The pKa of a silanol group is approximately 3.8, therefore, reducing the mobile phase pH will protonate these groups effectively disrupting the interaction with basic analytes and improving peak shape (Figure 22).   

 

Figure 21: Interaction of basic amphetamine molecule with a surface silanol group.  Amphetamine pKa 9.8, silanol pKa 3.8.

 
 

Figure 22:  Analysis of a sample mixture containing amphetamine (peak 7) at pH 7 (left) and pH 2.5 (right).

 
 

Modern approaches to working with ionisable analytes often involve working at extremes of pH to avoid variations in selectivity due to changes in the mobile phase pH.  The pH is typically altered using TFA, formic acid, ammonia, and ammonium hydroxide.  This often improves method robustness but requires selectivity to be optimized by other means, such as, stationary phase, organic modifier type, and eluotropic strength which limits the extent to which separations can be optimized.  For complex separations (acids/bases or mixtures of both) this approach will not always work and a knowledge of the pKa vs. eluent pH is critical.

A sample containing a mixture of acids, bases, and neutrals with a range of LogP (D) values has been analyzed using a similar approach to that outlined above.  The pH was adjusted to 2.2 by addition of 0.1% TFA.  The resulting chromatogram (Figure 23) shows an unsatisfactory separation.  Using modelling software for this separation (Figure 23) results in a highly complex resolution map which demonstrates that the pKa of the analytes and the pH of the eluent must be considered and furthermore investigated.  It should be noted that the pH value of 2.2 at which this separation was carried out is at a minima on the Drylab resolution map which will give poor resolution.

 

Figure 23: Analysis of a mixture of acids, bases, and neutrals with a range of LogP (D) values at pH 2.2.  Experimental chromatogram produced at pH 2.2 (top) and Drylab simulation over a range of pH values (bottom).

 
 

Using the Drylab simulation the pH for this complex separation can be optimized.  Figure 24 shows the chromatogram that was produced at the slightly increased pH of 3.2, in which all peaks are well resolved from each other.  Assessment of the resolution map shows that pH 3.2 is at a maximum and will, therefore, give optimum resolution.  It should also be noted that small changes in pH should not have a large impact on the resolution as predicted by the simulated resolution map.

 

Figure 24: Analysis of a mixture of acids, bases, and neutrals with a range of LogP (D) values at pH 3.2.  Experimental chromatogram produced at pH 3.2 (top) and Drylab simulation over a range of pH values (bottom).

 
 

Buffers

A buffer is a weak acid or base in co-solution with its conjugate acid/base.  It resists small changes in pH that may occur within the HPLC system.  A particular buffer is only reliable at 1 pH unit either side of its pKa (Table 10).  The buffer concentration must be adequate but not excessive.  Below 10 mM buffers have very little buffering capacity, therefore, they will not be able to resist changes in pH.  At concentrations greater than 50 mM there is a risk of the salt being precipitated in the presence of high organic concentrations (i.e. > 60% MeCN).  Buffer concentrations will normally be in the range 25 – 100 mM, and it is recommended that the effect of the buffer concentration is investigated as part of the method development process.  Remember that if a UV based detector is being used to take note of the UV cut off of the buffer that is being used.

 

Table 10: Properties of common HPLC buffers.

 
 

Again with LC-MS applications there are special considerations that must be taken into account when choosing a buffer.  Volatile buffers are required to avoid fouling of the API interface.  TFA is not a buffer and has no useful buffering capacity in the pH range usually associated with reversed-phase HPLC. Instead it is used to adjust the mobile phase pH well away from the pKa of the analytes such that small changes in pH that occur will not affect the chromatographic retention or selectivity.  However, we have seen previously that this approach is often not feasible and produces poor chromatographic results.  Furthermore, a significant disadvantage of TFA is its ion-pairing capability and its tendency to ion pair with ionized analyte molecules in the gas phase within the API interface and potentially drastically reduce MS sensitivity for certain analytes. TFA is best avoided unless one knows something about the interaction of TFA with the analytes under investigation.
Formic acid can be used in preference to TFA, for while it has ion-pairing capability, the ion pair strength is low enough such that when the associated pair move from the condensed phase into the gas phase within the API interface, the ion pair dissociates, allowing the gas phase charged analyte to be successfully detected by the mass spectrometer.  Often an increase in buffer strength will decrease sensitivity so this parameter should be optimized.

 

Some acids, bases, and buffers that can be used for LC-MS applications are listed below.

Acetic Acid

Formic Acid

 

Proton Donors

     

Ammonium Hydroxide

Ammonia Solutions

 

Proton Acceptors

     

Trichloroacetic Acid (< 0.02% v/v)

Trifluoroacetic Acid (< 0.02% v/v)

Triethylamine (< 0.02% v/v)

Trimethylamine (< 0.02% v/v)

 

Chromatographic Separation Ion-Pair Reagents

     

Ammonium Acetate

Ammonium Formate

 

Buffers

 
 

Analytical Chemists

  • I feel empowered to fix things
  • I can troubleshoot effectively
  • I know where to go for help
  • I understand my analyses
  • I know where to get applications
  • I’m up to date
  • I’m more employable
  • My career is progressing
 

Laboratory Managers

  • Improved equipment utilization
  • Faster method development/problem solving
  • Flexible workforce with a common standard
  • Better quality data
  • Get up to speed quicker
  • Lower T&E
  • Less reliant on me
  • I spend less time on training
 

Subscribe for $399 per/year and access:

  • The entire e-Learning archive
  • All Essential Guide Webcasts and Tutorials
  • LCGCs archive of articles and webcasts
  • Expert troubleshooting advice when needed
subscribe now
 
 


Like CHROMacademy on Facebook and keep up-to-date with the latest webcast, tutorial and eLearning release schedules.

 

Facebook

Stationary phase selectivity is critical for a successful (optimized) separation.  Separation on significant differences in LogP (D) (hydrophobicity) using a C18 column is not always successful.  Mechanisms to further discriminate between function group chemistry, shape (hydrodynamic volume), dipole (induced dipole) capability are all very important in modern HPLC.  This may involve the use of ligand chemistry (PFP and phenyl phases are becoming increasingly important), polar end capping, Type II and III silica, ligand density on the silica surface, polar embedded ligands etc.  In modern HPLC an awareness of the various column properties which are described in modern databases and by manufacturers is a must (further learning modules and webcasts on this subject can be found in CHROMacademy).

Screening a column for ‘selectivity’, selecting a column which is similar to or radically different (orthogonal) from one that has been used in the past can be a complex process.  Apart from experience and the literature, there are various in-silico column classification databases which use test results based on various chemical probes, which are designed to highlight specific column characteristics.  For example, the Tanaka test compounds can be used to compare the quality and performance of HPLC columns.  Results of these test probes can be graphically represented (Figure 25); the more symmetrical the hexagon appears and the larger its area the more balanced the stationary phase is in the sum of its chromatographic properties.

 
 

A: k (Pentyl benzene)

B: α (Pentyl/Butyl benzene)

C: α (Triphenylene/o-Terphenyl)

D: α (Caffeine/Phenol)

E: α (Benzylamine/Phenol; pH 7.6)

F: α (Benzylamine/Phenol; pH 2.7)

 

9.59

1.51

1.63

0.44

0.23

0.02

 

Figure 25: Radar plot of the results from the Tanaka test probes.

 
 

Links to Commercially Available Databases

Tanaka/Euerby & Peterson Column Selector Database
Column selection database based on the work of Euerby and Petersson:
//www.acdlabs.com/products/adh/chrom/chromproc/index.php#colsel


Note: in order to run this application you will also need to install the Freeware version of ACD Labs ChemSketch which can be downloaded here:
//www.acdlabs.com/resources/freeware/chemsketch/

 

USP & PQRI Databases

Both the USP and PQRI databases can be found at the following location:
//www.usp.org/app/USPNF/columns.html

 

Selectivity is the method developer’s friend.  During method development phases should be screened for orthogonality.  Even columns containing the same bonded phase i.e. C18, C8 etc., from the same manufacturer can show different selectivity (Figure 26).  The reproducibility of a column should also be investigated to determine if the separation will be susceptible to underlying manufacturing characteristics (silanol activity, metal ion content etc.).  Method development should always be carried out with a new column.

 

Figure 26: Herbicide analysis using three different Zorbax C8 columns.

 
 

When developing a method it is useful to have a simple measure of the column resolving power.  A very practical measure of the relative separating power of columns of different dimension is the so-called Resolution Capacity which is given by dividing the column length (mm) by particle size (μm) (Table 11).  When transferring methods resolution capacity needs to be conserved.  As can be seen in table 11 a reduction in column length accompanied by a reduction in particle size will result in a column with the same resolving power.  A gain in resolving power can also be achieved by increasing the column length while keeping the particle size consistent.

 

Table 11: Resolution Capacity of standard HPLC column geometries.

 
 

Analytical Chemists

  • I feel empowered to fix things
  • I can troubleshoot effectively
  • I know where to go for help
  • I understand my analyses
  • I know where to get applications
  • I’m up to date
  • I’m more employable
  • My career is progressing
 

Laboratory Managers

  • Improved equipment utilization
  • Faster method development/problem solving
  • Flexible workforce with a common standard
  • Better quality data
  • Get up to speed quicker
  • Lower T&E
  • Less reliant on me
  • I spend less time on training
 

Subscribe for $399 per/year and access:

  • The entire e-Learning archive
  • All Essential Guide Webcasts and Tutorials
  • LCGCs archive of articles and webcasts
  • Expert troubleshooting advice when needed
subscribe now
 
 


Like CHROMacademy on Facebook and keep up-to-date with the latest webcast, tutorial and eLearning release schedules.

 

Facebook

Parameters to Optimize

Several parameters are required to specify and optimize a gradient in a reversed phase HPLC method (Figure 27).

Initial %B – starting mobile phase composition (described in terms of the % of the strong solvent ‘B’).

Isocratic hold – a period within the gradient in which the eluent composition is held at the Initial %B. This achieves a degree of analyte focusing but also crucially enables easy transfer of gradients between different instruments based on the specific instrument Gradient Dwell Volume (VD).

Gradient time (tg) – the time during which the eluent composition is changing.

Final %B – final mobile phase composition.

Purging – usually achieved using a short ballistic gradient ramp to high %B in order to elute highly retained components (of no analytical interest) from the column. There may be an isocratic hold at this composition to ensure elution of all analytes and strongly absorbed components of no analytical interest.

Conditioning – returning the system (specifically the column) to the initial gradient composition. In practice, with modern instruments, this step is programmed to occur very rapidly.

Equilibration – the time taken to ensure the whole of the analytical column is returned to initial gradient composition. This is an important step and if not properly considered can lead to retention time and quantitative variability.

 

Figure 27: Typical gradient profile.

 
 

Of particular importance are the re-equilibration time and the gradient time.  The time required to completely re-equilibrate the analytical column prior to the next injection is dependent upon the column dimensions and the flow-rate used. Most manufacturers recommend passing through ten column volumes of eluent at the gradient starting composition for complete re-equilibration, however this can be determined empirically by shortening or lengthening the re-equilibration time and carefully observing any irreproducibility in retention times on successive injections of a test mixture. Further, the equilibration time may be shortened by increasing the eluent flow-rate (take care not to exceed the maximum system operating pressure) during the equilibration phase, but care must be taken to ensure pressure stabilisation at the original flow-rate prior to injection of the next sample.

 

The column volume can be calculated using equations 10 or 11.

 

Following manufacturers specification will result in a re-equilibration time, at the initial gradient eluent composition, PLUS the system Dwell Volume (see CHROMacademy for calculation of this).

 
 

Optimisation of Gradient Time and k*

Gradient and isocratic separations work differently – the separation mechanisms differ greatly between the two forms of chromatography.  There are sometimes examples where gradient methods are not sufficiently reproducible or where equipment struggles to form a gradient at high percentages of acetonitrile or when running ‘quick’ gradient methods.  There are several factors that can be optimized to improve the robustness of gradient methods and some quick calculations detailed below that can be used to check we’re getting it right.

The gradient that is being run must be fit for purpose; this can be checked using the following simple equation:

 

 

Where:
tg = gradient time (minutes)
F = flow rate (mL/min)
Φ = change in eluent composition (i.e. 0.4 for a 20 to 60% B gradient)
Vm = interstitial volume of the column, which is estimated by

 

 

S = shape selectivity factor which can be estimated by:

 

 

For analytes < 1000Da a value of 5 is typically used for S

k* = gradient retention co-efficient, k* is used opposed to k (for isocratic HPLC) because in gradient HPLC the retention factor of each analyte is constantly changing as the elutropic strength of the mobile phase is altered.

For a ‘good’ method ideally k* would lie in the range 2 to 10, for an ‘acceptable’ method k* should certainly be in the range 1 to 20.

If k* is too low, then there is a risk of interference from other sample components or analytes as the analyte does not have enough affinity for the stationary phase to differentially partition away from other sample components.  When k* is too high, the analysis time is unnecessarily long.

If a ‘fast method’ and a traditional method are considered values for k* can be calculated and assessed to see if they fall within the good or acceptable ranges to give a guide on the expected robustness of the methods.

 
 

Traditional HPLC method:

Column: C18 150 x 4.6 mm, 5 μm
Flow: 1.5 mL/min
Gradient: 20 to 65% Acetonitrile (0.1% Formic acid) in 7 minutes

 

 

A value of 2.75 is in the acceptable range indicating that this gradient is performing well! The gradient retention factor is over two, so problems associated with low retention would not be expected, and the value is not excessively high, therefore the analysis time will not be longer than required.

 
 

Fast HPLC method:

Column: C18 50 x 2.1 mm, 1.8 μm
Flow: 0.9 mL/min
Gradient:20 to 65% Acetonitrile (0.1% Formic acid) in 2 minutes

 

This gradient is also within the ‘good’ range of 2 to 10.  It may be possible to run the gradient a little faster without suffering too much from reproducibility problems.

As an aside the value of Vm can be used when estimating the re-equilibration time for the columns. Using the often recommended 10 x column volume for re-equilibration, the traditional column would take 11.3 minutes to re-equilibrate and the fast column just 1.3 minutes to re-equilibrate.

Of course  the equations above can also be used to predict the ‘ideal’ gradient time, depending upon the column and analysis speed requirements.

For a "fast" HPLC method, a 50 x 2.1 mm, 1.8 μm column might be used and the usual optimum flow rate for this column is usually around 600 μL/min. If a scouting gradient is required, this would typically be 5 to 95% organic.  A "fast" gradient would typically result in a k* value of 2, which allows a calculation of the optimum gradient time:

 

 

Such a fast gradient over a wide range of organic will require a very high performing pump and mixing system and as such specialist equipment will almost certainly be required.

The reproducibility of gradients will differ according to the manufacturing characteristics of the various instruments used.  UHPLC systems will have very low mixing volumes, low gradient dwell volumes and low extra column volume.  In more traditional HPLC systems, all of these volumes will be larger.

All of these factors will be responsible for the actual column content and the programmed gradient to be different.  Figure 28 shows a separation and the actual organic content at the column OUTLET and the programmed gradient overlaid. 12  It is more conventional to think about dwell time/volume at the column inlet, however, the principle holds true.

Figure 28 shows the magnitude of the differences between the programmed gradient and the actual composition of the eluent within the column at any point in time, which result from both gradient dwell volume as well as gradient mixing/forming issues (cavitation etc.).  Importantly any irreproducibility in the formation of the gradient, due to instrumentation problems, may result in retention time irreproducibility and possible changes in selectivity.

A good rule of thumb when considering gradient reproducibility is to keep the volume of the gradient at least double that of the gradient dwell volume. 13 

 

 

Where:
Vg = gradient volume (mL) (tg x F)
Vd = gradient dwell volume (mL)

 

In the fast HPLC example above the gradient volume is 1.08 mL (1.8 min x 0.6 mL/min). Therefore, the maximum "allowable" dwell volume for the instrument would be 540 μl.

 

Figure 28: Simulated chromatogram with overlaid gradient (dotted line). 12  The Y-axis represents the relative peak height and the actual modifier concentration at the end of the column as indicated by the solid curve.

 
 

Peak Capacity

The separation efficiency of columns under isocratic elution conditions is measured in theoretical plates.  Peak capacity is used to describe the separation efficiency for gradient elution.  Peak capacity describes the maximum theoretical number of components that can be successfully separated with a given column and set of analytical conditions with Rs =1 (Figure 29 and Equation 17). 11  Equation 17 can be used to give an approximation of the number of components that can be separated under a specific set of conditions, if this number is lower than the number of components in a sample than the method will not produce a chromatogram with resolved peaks.

Where:

tg = gradient time
w = average peak width at 4σ (13.4% of peak height)

Note: This is an approximation but a good guide.  The average peak width can be calculated by adding the peak widths of the first and last peaks and dividing by 2.

 

Figure 29: Calculation of peak capacity.

 

Peak capacity is a function of gradient time, flow rate, column length, and particle size.  Increasing column length while keeping particle size and gradient time constant results in a maximum value of peak capacity being reached, and in fact, for longer columns the value of peak capacity may decrease (Figure 30).  Improving peak capacity using particle size seems to give more promising results, with the decrease in particle size giving higher peak capacity values.  Increasing the gradient duration will increase the peak capacity; however, for longer gradients the increase in peak capacity with time becomes small as a maximum will be reached.  Peak capacity can be optimized using the flow rate at a fixed gradient time (tg).  Peak capacity will increase proportionally to the square root of column efficiency (Equation 18), therefore, doubling column efficiency will increase peak capacity, but only by 40%.

 

Where:

N = efficiency
B = slope of the function ln k (k is the retention factor) versus solvent composition C
ΔC = change in solvent composition
t0 = breakthrough time
tg = gradient time

 

Figure 30: Plot of tg / t0 vs. P giving the optimal peak capacity for a separation.

 
 

Analytical Chemists

  • I feel empowered to fix things
  • I can troubleshoot effectively
  • I know where to go for help
  • I understand my analyses
  • I know where to get applications
  • I’m up to date
  • I’m more employable
  • My career is progressing
 

Laboratory Managers

  • Improved equipment utilization
  • Faster method development/problem solving
  • Flexible workforce with a common standard
  • Better quality data
  • Get up to speed quicker
  • Lower T&E
  • Less reliant on me
  • I spend less time on training
 

Subscribe for $399 per/year and access:

  • The entire e-Learning archive
  • All Essential Guide Webcasts and Tutorials
  • LCGCs archive of articles and webcasts
  • Expert troubleshooting advice when needed
subscribe now
 
 


Like CHROMacademy on Facebook and keep up-to-date with the latest webcast, tutorial and eLearning release schedules.

 

Facebook

The Bandwidth parameter in Diode Array detection is related to the number of diode responses which are averaged in order to obtain a signal at a particular wavelength (Figure 31).  A wide bandwidth has the advantage of reducing noise by averaging over a greater diode range.  Noise is random; therefore, averaging the response over a large range of diodes will reduce noise.  As the bandwidth in increased, the signal intensity (detector sensitivity) increases as some diodes will result in a lower absorbance compared to a reading using only the single most intensive wavelength (λmax).  A wide bandwidth results in a larger range of wavelengths being averaged when producing a spectral data point, which results in a loss of spectral resolution.

A narrow slit width provides improved spectral resolution for analytes which give UV spectra with enough fine detail to be useful for qualitative analysis.  For example, improved spectral resolution will increase the confidence of library matching search results when attempting to identify unknown peaks within a chromatogram.  A wide slit width allows more of the light passing through the flow cell to reach the photodiode array, hence, the signal intensity and detector sensitivity will increase.  Baseline noise will also be reduced leading to an increase in signal to noise ratio.  However, with a wider slit width the optical resolution of the spectrophotometer (its ability to distinguish between different wavelengths) diminishes.  The wavelength of light falling on each diode becomes less specific as the light becomes more diffuse.  Any photodiode receives light within a range of wavelengths determined by the slit width, and so spectral resolution decreases.

The effect of changing band- and slit width are summarised in Tables 12 and 13.

 

Content on this page requires a newer version of Adobe Flash Player.

Get Adobe Flash player

 

Content on this page requires a newer version of Adobe Flash Player.

Get Adobe Flash player

Figure 31: Diode array detector.  Setting bandwidth and slit width.

 
 

Table 12: Summary of the effect of changing slit width on baseline noise and spectral resolution.

 
 

Table 13: Summary of the effect of changing bandwidth on S/N noise and spectral resolution.

 
 

Choosing Sample and Reference Settings

A diode array detector can compute and store several signals simultaneously and also manipulate the signals together in order to yield a composite or deconvoluted chromatogram.  The following signals are usually collected using diode array detectors:

  • Sample wavelength – the centre of a wavelength band with the width of the sample bandwidth
  • Reference wavelength – the centre of a wavelength band with the width of the reference bandwidth

The signals comprise a series of data points over time with the average absorbance in the sample wavelength band minus the average absorbance of the reference wavelength band.  An empirical method is detailed below and shown in Figure 32 for setting typical sample and reference wavelengths.

 
 

Step 1
The sample signal should be chosen at the lambda max (λmax) provided there will be no interference from the mobile phase absorbance.

Step 2
The sample signal bandwidth can be selected by imagining that the spectrum around λmax has a Gaussian peak shape.  Choose a bandwidth that is representative of the full width at half maximum.
Note: Not all peaks obtained using UV detection will be perfectly Gaussian, therefore, it can be difficult to choose a bandwidth.  The use of a bandwidth of 30 nm should suffice under these circumstances.

Step 3
The reference wavelength is chosen in an area of low or no absorbance as close to the sample signal as possible.  Select a wavelength at which the analyte absorbance drops below 1 mAu (e.g. 280 nm in Figure 26).  Add 10 nm safety margin (e.g. 290 nm) to give the start of the reference window.

Step 4
The reference wavelength should be set in the middle of a 100 nm window (if possible).
In the example shown (Figure 26) the reference wavelength would be 290 nm + 50 nm = 340 nm with a reference bandwidth of 100 nm.

 

Figure 32: Sample and reference wavelength settings. 

 
 

The reference wavelength compensates for fluctuations in lamp intensity as well as changes in the absorbance/refractive index of the background (i.e. mobile phase) during gradient elution.  During gradient elution the composition of the eluent will change and, hence, so will its refractive index.  To compensate for the change in refractive index properties a reference wavelength should always be set otherwise drifting baselines will occur (Figure 33).  Noise will also be reduced as the reference wavelength is moved closer to the sample signal.  Without any reference measurement all noise and variability in lamp intensity is recorded within the signal.  When using a reference signal all lamp intensity and background (mobile phase) variability is subtracted out of the signal being measured.  The closer the reference wavelength is to the sample wavelength the more effectively these background deviations are catered for and the better the detector sensitivity.  However, the reference wavelength should not be selected too close to the analyte wavelength or the signal intensity may be seriously reduced.  Choice of a proper reference wavelength can reduce variability and drift in the chromatographic baseline resulting in better signal to noise performance.

 

Figure 33: Chromatogram with reference wavelength set (top) and without reference wavelength set (bottom).

 
 

Analytical Chemists

  • I feel empowered to fix things
  • I can troubleshoot effectively
  • I know where to go for help
  • I understand my analyses
  • I know where to get applications
  • I’m up to date
  • I’m more employable
  • My career is progressing
 

Laboratory Managers

  • Improved equipment utilization
  • Faster method development/problem solving
  • Flexible workforce with a common standard
  • Better quality data
  • Get up to speed quicker
  • Lower T&E
  • Less reliant on me
  • I spend less time on training
 

Subscribe for $399 per/year and access:

  • The entire e-Learning archive
  • All Essential Guide Webcasts and Tutorials
  • LCGCs archive of articles and webcasts
  • Expert troubleshooting advice when needed
subscribe now
 
 


Like CHROMacademy on Facebook and keep up-to-date with the latest webcast, tutorial and eLearning release schedules.

 

Facebook

Electrospray ionization (ESI) and atmospheric pressure chemical ionization (APCI) are complementary ionization techniques as:

  • ESI used for medium to high polarity analytes
  • APCI used for medium to low polarity analytes

Electrospray ionization is highly compatible with analytes possessing the following characteristics:

  • Moderate to highly polarity
  • Up to 100 000 Dalton
  • Ionize in solution perhaps with multiple charges (z > 1)

Electrospray ionization favours the analyte in the ionized form; this is, they should live in the eluent solution prior to introduction into the API interface, although this is not an absolute pre-requisite for generating a response in LC-MS, as there are secondary methods of analyte charging within the ESI interface.

APCI finds most of its applications in molecular weights below 1000 Da for medium to low polarity molecules.  The analytes will need some degree of volatility and should not be thermo labile.

Under certain circumstances, the correct ionization mode is not immediately obvious (for example when dealing with moderately polar molecules); in such occasions, the following steps should be considered:

  • Infuse sample and evaluate ESI signal response (for both positive and negative ion modes)
  • Infuse sample and evaluate APCI signal response (for both positive and negative ion modes)
  • Choose the best ionization technique (consider signal response –intensity, linearity and time variability)
  • Optimize conditions (see below)

By infusing your sample and comparing signal response (in terms of intensity, linearity and time variability) obtained under different ionization techniques, it is possible to select the best choice before proceeding with the optimization process.  The process can be achieved by switching either polarity (positive or negative ion mode) or ionization technique (ESI, APCI, APPI…).

 
 

Real World Example

Dinitropyrenes are among the most mutagenic substances tested and have been classified as possible human carcinogens by the International Agency for Research on Cancer.3  Dinitropyrenes can be found in the emissions from diesel exhausts, gasoline engine emissions, and airplane emissions, to name but a few.  Emissions from diesel engines are the only source that pose significant risk to human health.  Previously GC-MS techniques have been used for the analysis of dinotropyrenes from plasma.  Straube et al. have analyzed dinitropyrenes and their metabolites, aminonitropyrenes and diaminopyrenes (Figure 34), from diesel engine emissions in plasma by HPLC-MS/MS to create a method for assessing human exposure. 

For the determination of LODs and LOQs of each of the analytes all ionization modes were assessed, electrospray ionization (ESI), atmospheric pressure chemical ionization (APCI), and atmospheric pressure photoionization (APPI) (Table 14).  As can be seen the highest sensitivity for the measurement of aminonitropyrene and dinitropyrene was achieved with normal phase HPLC-MS/MS using APPI ionization. 

 

Figure 34: Dinitropyrene (DNP) (1), aminonitropyrene (ANP) (2), and diaminopyrene (DAP) (3).

 

Table 14: Comparison of LODs [pg/10 μL] (S/N in brackets) and precision calculated on the basis of 20 replicate measurements of one sample within one day (concentration [pg/10 μL] in brackets) for reversed phase (RP) or normal phase (NP) HPLC connected to ESI, APCI, and APPI.2  * =  not performed. 

 
 

Optimisation: Eluent & Gas Flow Rates

Once the correct ionization mode has been selected there are parameters that should be optimized, values of which are summarized in Table 15.

 

Table 15: Optimisation of parameters in LC-MS.

 
 

In ESI the threshold electrospray voltage (also called the onset voltage or VON) is the applied voltage which destabilizes the Taylor cone (which is formed at the capillary tip) and initiates the ion evaporation process.  This voltage can be estimated using Equation 19.  For the most part users only have full control over the solvent surface tension parameter.  In reversed phase HPLC the percentage of aqueous and organic solvents determines the surface tension of the eluent.  From Equation 19 at a given electrospray voltage greater than VON a higher organic content in the mobile phase will result in more rapid and complete desolvation which gives more efficient ion evaporation at the interface and an  improved MS signal.  If the surface tension of the eluent is higher, a higher threshold voltage will be required to initiate the ion evaporation process (Table 16).

 

 

Where:

γ = solvent surface tension
rc = LC capillary outer radius
d = distance from the LC capillary tip to the MS inlet

 

Table 16: Threshold electrospray voltages. 4

 
 

ESI of Parabens

A mixture of C1 – C4 parabens (100 ppm of each) were analyzed by reversed phase HPLC on a C18 column with acetonitrile/water.  UV and ESI MS detection were both used.  The chromatogram obtained from HPLC-UV shows good detection of each of the parabens (Figure 29).  However, using ESI-MS in negative ion mode illustrates that the larger hydrophobic chain increases the ionization efficiency (Figure 35). 

 

Figure 35: Structure of “paraben” (top).  Chromatogram obtained from the separation of a mixture of C1 – C4 parabens using HPLC with a UV detector (middle) and using HPLC with ESI MS (bottom).

 
 

How to Obtain Good ESI Results

There are several tricks of the trade that can be used to obtain good ESI results.

1.  Reduce salts (watch out for sodium adducts)

  • Use plastic vials
  • Eliminate soap and detergent

2.  Keep “leakage current” low

  • Reduce TFA and eliminate strong acids
  • Keep acid concentration low
  • Keep probe voltage low
  • Increase % of organic solvents, but always have some water in the mobile phase

A good rule of thumb with most MS parameters is that
If a little bit works, a little bit less probably works better
.

 

In MS LOD are controlled by

  • Mode of sample introduction
  • Efficiency of ionization
  • Efficiency of ion transmission
  • Efficiency of ion detection
  • “Noise” generated by system (electronic and sample)
  • Scan speed

Sensitivities are typically in the range 10-9 - 10-12 (-15) g.  If these limits of detection are not achieved pre-separation, extraction, or enrichment may be required.

 
 

Analytical Chemists

  • I feel empowered to fix things
  • I can troubleshoot effectively
  • I know where to go for help
  • I understand my analyses
  • I know where to get applications
  • I’m up to date
  • I’m more employable
  • My career is progressing
 

Laboratory Managers

  • Improved equipment utilization
  • Faster method development/problem solving
  • Flexible workforce with a common standard
  • Better quality data
  • Get up to speed quicker
  • Lower T&E
  • Less reliant on me
  • I spend less time on training
 

Subscribe for $399 per/year and access:

  • The entire e-Learning archive
  • All Essential Guide Webcasts and Tutorials
  • LCGCs archive of articles and webcasts
  • Expert troubleshooting advice when needed
subscribe now
 
 


Like CHROMacademy on Facebook and keep up-to-date with the latest webcast, tutorial and eLearning release schedules.

 

Facebook

Consistency of the integration method that is used is important.  A threshold value should be set (in mV) for peak identification.  If the threshold is too high small peaks will be missed, and conversely, if it is too low noise peaks will be detected.  S/N ratio should be at least 1:10.  If the number of peaks detected is strongly dependent on the threshold value the integration parameters are not rugged. 

The simplest way to determine the best integration technique is to run a set of known samples during or before validation and collect data for both peak height and area.  Calculate the results using both techniques and use the method that gives the most accurate and precise results.  It is also worthwhile to become familiar with the different types of integration and the errors that can occur.  Several review articles are available. 7-10  Some common integration errors are depicted in Figures 36 and 37.

 

Some guidelines to avoid error with integration of overlapping peaks are:

  1. Resolve the peaks!  The required resolution will depend on the ratio of peak heights.  A resolution value of 1.5 may be adequate for peaks that are of similar size, however, a higher resolution value may be needed if the peak area ratios are larger than 20:1.
  2. Using peak heights can be less inaccurate than peak area integration.  The use of peak height for quantitation is almost error-free, even in the presence of tailing peaks.
  3. Avoid tailing.  Tailing reduces resolution and will have a detrimental effect on integration.
 

Rule of Thumb
If the minor peak is <10% of the height of the major one, skimming the peak is the appropriate integration technique.
If the minor peak is >10% of the height of the major one, a perpendicular drop to the baseline connecting the true baseline before and after the peak group is best.

 
 

Figure 36: Common integration errors.  (a) Baseline before peak falsely identified, (b) improper integration of minor peak on the tail of a major one, and (c) wrong peak endpoint selected.  Solid integration baselines are drawn improperly, dashed lines show the correct integration. 5

 
 

Figure 37: Common integration errors. 6

 
 

Analytical Chemists

  • I feel empowered to fix things
  • I can troubleshoot effectively
  • I know where to go for help
  • I understand my analyses
  • I know where to get applications
  • I’m up to date
  • I’m more employable
  • My career is progressing
 

Laboratory Managers

  • Improved equipment utilization
  • Faster method development/problem solving
  • Flexible workforce with a common standard
  • Better quality data
  • Get up to speed quicker
  • Lower T&E
  • Less reliant on me
  • I spend less time on training
 

Subscribe for $399 per/year and access:

  • The entire e-Learning archive
  • All Essential Guide Webcasts and Tutorials
  • LCGCs archive of articles and webcasts
  • Expert troubleshooting advice when needed
subscribe now
 
 


Like CHROMacademy on Facebook and keep up-to-date with the latest webcast, tutorial and eLearning release schedules.

 

Facebook

  1. Dolan, J. W.; LCGC North America [Online] 2010, 28.

  2. Straube et al. J. Am. Soc. Mass Spectrom. 2004, 15, 1853-1862.

  3. Rosenkranz, H. S.; Mermelstein, R. Mutagenicity and Genotoxicity of Nitroarenes.  All Nitro-Containing Chemicals were not Created Equal. Mutat. Res. 1983, 114, 217–267.

  4. Klink, F.; //www.sepscience.com/docs/Bespoke/Editions/MSSolutions/MSSol7EU.pdf.

  5. Dolan, J. W.; LCGC North America: Integration Problems [Online] 2009.

  6. Meyer, V.R.; Solvent Properties.  In Practical High-Performance Liquid Chromatography, 4th Edition; Wiley-VCH Verlag GmbH & Co, KGaA, 2004; 79.

  7. Dyson, N.; Chromatographic Integration Methods, Royal Society of Chemistry, London, 2nd ed., 1998.

  8. Ogan, K.; In Quantitative Analysis using Chromatographic Techniques, Katz, E. (ed.), John Wiley & Sons, Chichester, 1987, p. 31.

  9. Rossi, D. T.; J. Chromatogr. Sci., 1998, 26, 101.

  10. E. Grushka and I. Zamir, in High Performance Liquid Chromatography, P. R. Brown and R. A. Hartwick (eds), John Wiley & Sons, New York, 1989, pp. 529-61.

  11. M. Gilar et al., J. Chromatogr. A 2004, 1061, 183-192.

  12. G. Hendriks, J.P. Franke, and D.R.A. Uges, J. Chromatogr. A,1089 (2005) 193.

  13. Dolan, J.W. LCGC North America, August 1, 2011.

  14. Dolan, J.W., and Snyder, L.R.; Troubleshooting LC Systems, Humana Press, Clifton, 1989, p. 85.

  15. M. Z. El Fallah, M. Martin, Chromatographia 1987, 24, 115-122.
 
 

Analytical Chemists

  • I feel empowered to fix things
  • I can troubleshoot effectively
  • I know where to go for help
  • I understand my analyses
  • I know where to get applications
  • I’m up to date
  • I’m more employable
  • My career is progressing
 

Laboratory Managers

  • Improved equipment utilization
  • Faster method development/problem solving
  • Flexible workforce with a common standard
  • Better quality data
  • Get up to speed quicker
  • Lower T&E
  • Less reliant on me
  • I spend less time on training
 

Subscribe for $399 per/year and access:

  • The entire e-Learning archive
  • All Essential Guide Webcasts and Tutorials
  • LCGCs archive of articles and webcasts
  • Expert troubleshooting advice when needed
subscribe now
 
 


Like CHROMacademy on Facebook and keep up-to-date with the latest webcast, tutorial and eLearning release schedules.

 

Facebook

Learn chromatography from the experts

Whether you work in a lab or manage a lab, you will benefit from being a member of CHROMacademy.

As a member of CHROMacademy, you will get access to our vast e-Learning archive full of great interactive content and animations.
All our Essential Guide Webcasts and tutorials and LCGCs archive of magazine articles and webcasts from your favourite authors - John Dolan, John Hinshaw, Mike Balough, and Ron Majors. Plus vendor application notes, electronic laboratory tools and calculators and with our 'Ask the Expert' function - help is always at hand.

 

Analytical Chemists

  • I feel empowered to fix things
  • I can troubleshoot effectively
  • I know where to go for help
  • I understand my analyses
  • I know where to get applications
  • I’m up to date
  • I’m more employable
  • My career is progressing
 

Laboratory Managers

  • Improved equipment utilization
  • Faster method development/problem solving
  • Flexible workforce with a common standard
  • Better quality data
  • Get up to speed quicker
  • Lower T&E
  • Less reliant on me
  • I spend less time on training
 

Subscribe for $399 per/year and access:

  • The entire e-Learning archive
  • All Essential Guide Webcasts and Tutorials
  • LCGCs archive of articles and webcasts
  • Expert troubleshooting advice when needed
subscribe now

 


Like CHROMacademy on Facebook and keep up-to-date with the latest webcast, tutorial and eLearning release schedules.

 

Facebook

loading data
loading data
loading data
loading data
loading data
Home | About UsContact Us | SubscribeTerms and Conditions | Advertise | Privacy Policy |

loading data

loading data

loading data

 

loading data


loading data

Where do you start when you want to develop a better HPLC method? Whether you work in a regulated environment or not, setting specifications for your HPLC protocols is a good practice. In this Essential Guide we will set out guidelines for chromatographic parameters such as retention, resolution, and efficiency that can be used to improve your HPLC methods prior to method validation. We will explain how mobile phase design and instrument parameters can be used to achieve improvements in method robustness and what to look for when a method is failing. Practical tips for sample and eluent preparation, and the correct detector settings to use will also be discussed.

Kevin A. Schug
Associate Professor
University of Texas, Arlington

Dawn Watson

Dawn Watson
CHROMacademy Technical Expert
Crawford Scientific

Key Learning Objectives:

  • Understand what specifications should be in place so that HPLC methods are robust and reproducible
  • Learn good practice for sample preparation and injection
  • Understand the importance of using the correct detector settings for your analysis
  • Appreciate the variables in LC(MS) methods which effect robustness
  • Develop knowledge of how to alter retention and selectivity of analytes
  • Gain practical tips to improve your method development process