The main ERROR sources are

1.  Incomplete separation, peak overlapping because of non optimized selectivity or
     insufficient separation power.
2.  Detector non linearity or working outside the detector working range.
3.  False calibration.
4.  Poor integration because of wrong base line adjustment or too low peak-start /
     peak end detection sensitivity and wrong “separation line positions”..

HPLC made in recent years a quantum jump in separation power since the development of “comprehensive HPLC” or LC x LC. Another important improvement is “chip LC”. Monolithic columns and short separation systems with smaller than 2 micrometer particles improve drastically the separation speed and thus the quantitative analysis of substances in low concentration. Mobile phase pre heaters allow for gradient free higher temperature separations up to even 200 degree centigrade, also a step forward. Meanwhile the data rate of detector signal storage has been lifted from the classical 10 Hz up to 100 Hz making fast LC qualified for correct integration. The development of more stable stationary phases continues all the time. Precisely temperature controlled fluorescence detection allows signal enhancement at low temperatures, but still HPLC detectors need (or have) quite a large volume (10 microliter) which is far away from super fast GC detector volumes. Although light pipe concepts help to improve the situation especially for the micro column application. The miniaturization of HPLC was for many years blocked because of the limitations of available detectors, but now we see a break through made with multi  parallel mini column systems, especially for the fast mass analysis in the biochemical research and production areas.  A serious limitation of quantitative HPLC is the relatively poor detectability limit in trace and purity analyses, as even HPLC/MS and -MS/MS systems could not do enough at trace concentrations below 0.1 weight-%. Now the new “Corona CAD” offers an improvement of more than one order of magnitude down to concentration ranges of 0.01 weight-% as it gets qualified signals still with 1 to 10 ng for certain substances. This however is far away from the quantitative analytical trace limits of GC (reaching levels below ppt = 0.001 ppb at femtogram detection limits) and it is also far away from the trace analysis level of fully instrumentalized HPTLC which also offers parallel mass analysis in the ppt concentration level. Drinking water impurity quantitation is an example. HPLC develops despite blocking regulation fast again. You should remain in contact with leading instrument producers, they also know everything about the still large series of international and national HPLC development and application conferences.
NOTE:  The recently developed new micro planar chromatographic technique µ-PLC (see in this SITE) allows to sample a full milliliter of HPLC eluate made with ONE PEAK onto a cleaned HPTLC plate, dries this huge sample amount at room temperature and focuses the residual material into a sharp circle, separates it into hopefully ONLY one substance and if not: you have a problem. The HPLC peak then is a substance mixture.


We invite HPLC-only chromatographers to have a look into the page “Quant. Error GC
and “Qual. Error PLC”. Many details mentioned under GC exist either exactly the same way in HPLC or are so similar that equal ways to reduce the error level are possible.

HPLC-only chromatographers should know, that the many years existing statement “GC is now over, we do it all by HPLC” was never correct and is since the break through of comprehensive capillary GC not anymore used. Comprehensive capillary GC produces data in the range of around 30,000 peaks per sample. Modern micro GC started in the year 2000. The first time micro miniaturized even temperature programmable process capillary micro gas chromatographs with about 5 detectors simultaneously inside 0.1 mm capillaries can separate mixtures without elution.
Miniaturized HPTLC separates multi dimensionally and is an unbeatable supporter for qualitatively uncertain HPLC separations. In this site nobody sees that the HPLC expert looks to such specialities. He has not to ask for permission in having a look to the chromatography “foreigners” but he may find a problem solution for himself when looking to the often only disqualified “others”

Help to reduce the errors under 1.:

Add-on of PLC as HPTLC or circular micro-PLC (PLCC) - see JPC, Journal of Planar Chromatography - Modern TLC, Springer, ISSN 0933-4173; see as an example: R.E.Kaiser, Methods of Detecting and / or Reducing Systematic Errors...., Part 2: Systematic Errors Caused by Separation Systems, J.Planar Chromatogr. 18 (2005) 118-126

It is a very simple task to transfer each peak in its mobile phase solution onto a 100 x 100 mm HPTLC plate using the same stationary phase as for the HPLC separation but a quite differing mobile phase is used. Or spray the peak solution onto a HPTLC plate with a completely differing stationary phase, focus the sample lines and use several mobile phases to reseparate. See in the above given reference about a quick micro procedure for phase selection with a mini amounts of even expensive phase not available for HPLC.

HPTLC offers many differing stationary phases. Very many more mobile phases are applicable as compared to what HPLC allows due to the used detectors, the applicable temperature, the economy. Micro-Circular planar liquid chromatography (PLCC) needs a few seconds for a separation of possibly mixed substances which have been eluted as one single peak in HPLC. Micro HPTLC needs only a few nanoliter sample, only a few microliter mobile phase and has far more than hundred detection modes available. Here a word of pre-information: don´t be shocked in case one sharp HPLC peak is PLC separated into several new ones. A changed identity cancels all former quantity values. And better you find it out than your chromatographer colleague.

We propose to use a type of “Trennzahl” instead of the theoretical plate height based plate number in HPLC. The majority of all HPLC analyses run phase programmed. Any programming makes the use of plate numbers and plate height values, and the complete “HETP / u curve”  chromatographically “illegal”. The latter is an often used concept to interpret the physical behavior of HPLC packings, packing procedures and the final characterization of a ready made HPLC column. But in the minute the separation is done under temperature programming (seldom), flow speed programming (seldom) and mobile phase programming (very often) the HETP / U curve concept and all other just mentioned column quality values are useless. The only one remaining is the Trennzahl but see “figure 5” in Qual.HPLC ERRORS for the real optimum under changed mobile phase flow speeds of figure 1 on this page below.

The TZ value is a way out for correct flow and temperature optimization in HPLC. The latter is still underdeveloped in HPLC but often very helpful. In order to get TZ values the chromatographer needs homologues. The group of di-nitro-benzoic acid esters is quite widely applicable and the esters can easily by synthesized or taken on the market. 

BUT: these are weaker ways to optimize a separation. And an optimized separation is the key to reduce possible quantity errors in HPLC.


Figure 1 below shows the direction of physical optimization in quantitative HPLC:
It is NOT the largest plate number or the smallest “height of a theoretical plate” (HETP) as too often stated in the literature. A small HETP value only tells about the physical quality of a column packing and - if well packed - about the quality of the packing material itself if in addition the pressure drop in the column is as small as possible. For a minimum danger for producing wrong quantitative date we need a high separation power at best expressed by the Trennzahl. MOST important is to select such a chemical selectivity, that the most critical pair of to be separated substances in the analysis have such a wide peak distance, that there is baseline separation. Only NOW we can optimize the separation speed: it should be so high, that the baseline separation around the substances in question is acceptable, the mobile phase back pressure is no problem for the life time of the HPLC pump and the PCreal number is as high as possible. This brings together data quality and economy. Figure 1 below shows, that with a column as mentioned earlier the Trennzahl and the real peak capacity (as number = PCreal) does drop only slightly with increasing phase flow speed. The analytical data productivity however rises drastically and thus isocratic HPLC should run at much higher flow speeds than proposed by classical theoretical concepts.

Figure 1 below shows how strong the fundamental ABT data change with the phase flow speed. As ALL chromatography values are based on “a, b0 and tm” - with other words all separation quality numbers can be calculated with “a, b0 and tm” - we understand, how dynamically is chromatography. It is time dependent. The residence time in the mobile and stationary phase controls everything. The mobile phase flow speed must be very constant in isocratic HPLC and must be very well reproducible in programmed HPLC. But also the programming values must be either highly constant or very well reproducible.

The reason for these mentioned conditions is easy to understand:

The peak height of a given substance depends on the mobile phase flow speed. The peak position depends on the peak residence time values tm and ts. tm changes with the flow speed, ts with temperature, flow speed, elution power of the mobile phase and retention power of the stationary phase. The peak height is a measure of the grams substance per ml mobile phase. This value is non linearly correlated with the substance concentration in the sample. In order to end up with a correct analytical result like weight-%, mole-%, gr per liter etc. we need a substance specific correction factor found by non linear calibration. If anything in HPLC changes the peak position we would need a “corrected” correction factor, that means a new calibration. But because the quantitative calibration function is non linear, we also need a very well reproducible and constant sampling process.

HPLC users should understand one critical statement given at a large international HPLC symposium and not rejected by any of the top experts: “Analytical data measured by (instrumentalized) PLC are less prone to quantitative errors than HPLC or HPLC/MS or -MS/MS data. The closest technical and methodological cooperation of HPLC with PLC (HPTLC) helps to find and reduce possible serious analytical errors qualitatively and quantitatively.”

This is the reason why we treat all chromatography modes using a uniform and as much as possible equal theoretical concept. Therefore we had to leave the classical basics.

The “error detector sf4” works also in HPLC.

With respect to the physical separation power and trace detectability HPLC is far behind GC, thus a combined use of capillary GC together with HPLC is powerful. HPLC acts in this case as sample preparation method.

With respect to the chemical separation power by the use of mobile phase programming, differing stationary phases and multi phase separation capabilities,  HPLC is reasonably far behind modern PLC / HPTLC, especially in its circular mode, thus a combined use of both is advisable.

With respect to the molecular weight range of applicability HPLC is unbeatable and its further development into packed capillary versions and into the electrical driven separation as well as in micro column switching techniques has a wide field of coming developments. We remind to a forgotten mode of physical polarity switching in micro HPLC, as it is working also with packed HPLC columns and offers best possibilities in trace analysis. Reason: not the mobile phase is programmed but the stationary phase effects are electronically changed by local temperature fields. There is no need to mechanically  “move an oven”. One can also change the temperature of two or more series coupled columns locally by electronics and thus have physical polarity programming controlled by software.

These are main reasons to reject method regulation. The latter means “stop development” but development is not only important. It is vital in modern bio chemistry analysis.


Figure 1: Strong and non linear changes of the fundamental data of chromatography: “a, b0, tm” with the mobile phase flow speed.
NOTE: The Trennzahl and the “PCreal” value, the number of baseline separated peaks within the chromatography range from k=0 up to k=10, does show nearly no change with the flow speed over a range from 0.2 mm/sec up to 2 mm/sec in a 250 mm long EUROSPHERE packed column. The Trennzahl maximum is reached at about
0.3 mm/sec flow speed, much too slow for an economical use of this column for mass routine analysis.

For facts and remarks to point 2, 3 and 4 click   “here”.

[Home] [We can help] [Systematic C-Errors] [Statistics] [Error Detector "sf4"] [Sampling/Calibration] [Qual.Error GC] [Quant.Error GC] [Qual.Error HPLC] [Quant.Error HPLC] [Qual.Error PLC] [Quant.Error PLC] [Integration] [Chrom. Combination] [µPLC Micro Planar LC] [Altern.Chrom.Theory] [Contact IfC] [About the Author]
[Home] [We can help] [Systematic C-Errors] [Statistics] [Error Detector "sf4"] [Sampling/Calibration] [Qual.Error GC] [Quant.Error GC] [Qual.Error HPLC] [Quant.Error HPLC] [Qual.Error PLC] [Quant.Error PLC] [Integration] [Chrom. Combination] [µPLC Micro Planar LC] [Altern.Chrom.Theory] [Contact IfC] [About the Author]