Translator for HPLC HINTS and TIPS for Chromatographers

Saturday, March 29, 2014

External (ESTD) vs. Internal Standard (ISTD) Calibration in HPLC

Reliable quantitation of sample analytes using HPLC analysis requires accurate and reliable quantitation of a standard(s). For chromatography applications, we commonly use either an External Standard or Internal Standard, as required, to insure reliable quantification of the sample.


  • NOTE: A quick comment about calibration methods. Before you begin to create any calibration tables or analyze any standards/samples, please make sure that your current chromatography method is reliable. It must retain the compound(s), be reproducible and resolve apart all of the samples and possible impurities with near to perfectly symmetrical peak shapes. Your calibration results will only be as good as your original method. A poor quality method may not provide reliable results so be sure and spend as much time as possible developing the initial HPLC method to be as rugged and reliable as possible before starting any quantitation or calibration. *Poor quality method development is the number one reason for problems with quantitation.


Methods of Quantitation, Peak-height vs. Peak-area: Both types of response provide a measurement of the detector signal. Proper and reproducible integration of the signals is critical. Peak area is the most popular choice in chromatography, but peak height measurements can also be used if the peaks have near perfect symmetry (desirable, but very rare). Whichever method you chose, you must use it consistently.

Definitions, External & Internal Standards: For most samples, there are two commonly used types of standards used. When known standards are run separately from the actual samples (in their own chromatogram) and their response is compared to that of the sample in another chromatogram, then we refer to this as an External Standard (ESTD). When the standard is added to the sample and analyzed at the same time we refer to this as an Internal Standard (ISTD). With an Internal Standard we are comparing the instrument response of the sample to a reference standard, both run together.

External Standard Calibration Notes: The sample must fall within a range bracketed by the calibration solution. I suggest that you include a range which covers concentration values which are ~ 50% or more outside of the expected range. Dissolve the final calibration standards into the mobile phase (or a weaker solution) when preparing the injection vials from the stock solution. At least five different concentration values should be used per order of magnitude (large range = more stds). *Inject the same volume of solution (different concentration) for each calibration standard (level) onto the column. Do Not inject different volumes of solution from one std vial. Plot peak response vs concentration. Make sure you have a linear response and that the line goes through the origin (zero intercept, ideally). Once you have injected all of the standards, repeat the process again at least three more times (or use multiple injections) to determine overall reproducibility before constructing the final calibration table.

Internal Standard Calibration Notes: Internal standards are commonly used when many sample preparation steps are required before the sample can be injected onto the column. The internal standard may compensate for any losses during filtration or extraction. Selection of the Internal Standard is critical. Some of the characteristics should include: It must be different than the sample and must not elute where any sample peaks could be expected; It should not elute where any interfering matrix or other compounds could appear; It should have a similar linear response as the sample (Inject a fixed volume/concentration); Available in a high purity form from one or more commercial sources (certified method); Must be stable and not react with the sample or mobile phase solution. 

Add it to the samples before any extraction procedures. Base the amount of ISTD concentration such that it is between 1/3 and 1/2 of the expected concentration of the sample(s). The sample's target concentration range is a good value to use. *Because of these and other strict conditions, finding a suitable Internal Standard can take some time and testing.

24 comments:

  1. Why not use different injection volumes from the same solution if this prepared accurately?

    ReplyDelete
    Replies
    1. An excellent question. Generally, injecting different volumes from the same vial to create a calibration curve is an unacceptable procedure in chromatography. There are many reasons for this. One reason is that autoinjectors are usually more accurate at injecting the same volumes over-and-over reproducibly. If the volume of your stds and samples are the same (i.g. 20 ul), then they will all maintain the same accuracy of volume (no matter what it actually is). Another reason is linearity if often better with stds at the same volume. If you were to inject 1, 2, 5, 10, 20, 50, 75, 100 ul from the same vial and use that data for your calibration curve, then the linearity of the resulting curve may be different. Each of those points may be off in actual volume. Using the same volume for all injections removes some variablility of the injector from the method resulting in a more reliable method. Lastly, another reason is that the injection volume directly effects retention time. Large volumes usually have longer retention times than small volumes so your peaks will have a wider, less accurate window to calibrate them.

      Delete
  2. How do you determine the injection volume of the internal standard?

    ReplyDelete
    Replies
    1. You should use the same injection volume for your samples and standards. This minimizes errors in Rt and area integration.

      Delete
  3. Hi, while I completely agree that you should always inject the same volume for a standard vs a sample in HPLC, I am having trouble convincing a co-worker that it is important. I have used all your arguments above but they believe that if the linearity of the HPLC in shown in regular calibrations of the instrument then it is fine. Are you aware of any regulatory (eg USP) references that I can use to convince them? Thanks Betty

    ReplyDelete
  4. All of the guidelines that I am aware of state that you are validating ONE set of analysis conditions so you can not use different injection volumes for the method as you please. If you inject different volumes of the same standard to form your Calibration Table, then the linearity of the injector is the only thing being tested, not your chromatography. Calibration Tables must be uniform and constructed using the same conditions as your samples will see. Your sample(s) and standards need to be the same volume. What happens when you calibrate with 5ul of std, but inject 20 ul of sample? The analysis method conditions are no longer the same so your method is out of compliance.

    ReplyDelete
    Replies
    1. Thanks for your input, I will continue to argue till the point is heard. Best Regards, Betty.

      Delete
    2. Another way of suggesting that you should always use the same volume is as follows. If you choose to use different volumes for standards/samples, than you will need to prove, document and validate that changing the volume does not change the results. Since this is not the normal way we do things, a good regulator should ask this question and require documentation/proof. On the other hand, using the same volume for standards and samples means you do not need to prove anything extra at all. The method used is the same each time. No extra work, no extra documentation, no issues with validation.

      Delete
  5. Hi,

    What is the concentration and volume should I choose for the Internal Standard if my calibration range is 50nmol/L to 1250nmol/L?

    Thanks.

    ReplyDelete
    Replies
    1. We normally select a concentration value such that it provides a signal value that is in-scale with your other samples. Typically, one that results in the same approximate area as your other stds and expected samples. If your calibration range has been correctly bracketed around the expected sample range, then a value at or near 50% usually works well. In your example this might be ~650 nmol/L or it might be a value that better approximates the area of your sample.

      Delete
  6. I’m a high school student and new to research & HPLC. I have a question.

    In using HPLC for chemical stability analysis. I was told to not compare the peak area of the compound each time but rather the ratio of peak area of the compound to peak area of the internal standard. But I don’t understand why is that, why don't we just compare the actual peak area of the compound?Why did we use the “ratio of the compound to the internal standard" and not the peak area of the compound itself?

    ReplyDelete
    Replies
    1. Fatima: We use the ratio because the Internal Std's response will compensate for any losses or changes the same may have undergone during preparation,filtration, extraction, cleanup or analysis. *That is why an Internal Std is chosen for some methods instead of an External Standard.

      Delete
  7. Hi,
    When I create standard, should I add different compounds in same vial and set calibration levels for it?
    Alternately, is it possible to have different compounds in different vials and create a calibration curve with all these compounds in the curve?

    ReplyDelete
    Replies
    1. Several ways to do this. ESTD Example: You have 4 different standards at 5 different calibration levels. Place all 4 stds, at LEVEL 1 concentration in vial#1. Place all 4 stds, at LEVEL 2 concentration in vial#2. Repeat same procedure for Levels 3, 4 and 5. Each std can also be acquired at its own most appropriate wavelength (with a DAD). This is an example of a Multi-level calibration and we use them often in method development when we have different stds which need to be run on the same method.

      You could also place on std in one vial at one level and run them all individually. This will take more time, waste more solvent and generally be less reproducible than the above example (but it will work).

      Delete
  8. Hello, do I need to build a calibration curve every time I perform an analysis? Can I just keep the curve in the software and use it any time I need? Or is there a time when I need to build a new one?

    Thank you

    ReplyDelete
    Replies
    1. No. A calibration table is needed only when quantifying. The answer to if you need to "build" a new table each time depends on the software you are using. The calibration curve should be integrated into your analysis method (Most modern CDS have this built-in). Once the analysis method has been saved, it can be used for that specific sample again and again. *Note: You will still need to perform timely re-calibrations of the stds to verify the method.

      Delete
  9. Ok I understand. Thank you

    ReplyDelete
  10. How large of a range of external standards can I use? Typically I do a 500, 1000, 2000, 4000ppm set and get the samples to fall in that range (which isn't too hard). Should I make it narrower?

    ReplyDelete
    Replies
    1. Kyle: As with all calibrations, the range you select is determined by the range you wish to calibrate. The goal is to interpolate, never extrapolate so make sure the range you select covers the expected sample concentrations. Choose realistic values that are above and below the expected concentrations seen for the outside limits (high and low). Be sure and include plenty of values within these two limits, esp if more than one order of magnitude.

      Delete
  11. hi,
    i would like to ask if i inject my standard containing a number of compounds and an internal standard many times. Each time, the area of the compounds seems different between injections but also for teh internal standard. the ratio stay stable. could i use then the system for a calibration or i have to fix the problem regardless the similarity in ratio!!!?

    ReplyDelete
    Replies
    1. The internal standards job is to compensate for changes. It sounds like it may be doing just that for your application. The question I would ask is are the areas stable for the standards only? Injecting just the standards should give good repeatable results on their own. *If they do not provide acceptable results, then you need to fix the problem causing the poor results.

      Delete
  12. Thank you for your blog.

    ReplyDelete
  13. Hi,can u please explain how peak areas and peak nights can be calibrated by the intenal standard method

    ReplyDelete
    Replies
    1. Sure, this is covered in another one of my free articles; "Internal Standard (ISTD) HPLC Calculation Notes" [http://hplctips.blogspot.com/2016/02/internal-standard-istd-hplc-calculation.html].

      *Please use the Search feature on this blog to find related content. It is a great way to find additional information.

      Delete