I am getting ready to validate an HPLC method for the determination of a pharmaceutical drug on glass fiber filters. During robustness testing of flow rate, an odd observation was made. Keeping all other method constants the same, the flow rate was tested at 0.8, 1.0, and 1.2 mL/minute. The data show the following:

0.8 mL/min; Area= 1665, Height= 132; Width= 0.8 min, Rt= 12.1 min

1.0 mL/min; Area= 1338; Height=122, Width=0.7 min, Rt=9.8 min

1.2 mL/min; Area=1120; Height=114, Width=0.65 min, Rt=8.3 min

The observed trend is that as the flow rate increases, the peak area decreases (as does height and width of the peak). This does not make a lot of sense to me. Flow rate should effect the retention time obviously, but with all other constants being the same, it should not effect peak area at all. Any ideas about this observation?

Some other information: Waters LCM1 with UV detection at 230 nm being used with an XTerra RP-8 column. MPA = 5 mM NH4OAc, MPB = 5 mM NH4OAc/ACN (2/8), gradient elution. The drug is proprietary and I don't even know the structure of it. In addition, the method has been run several times at 1.0 and 1.2 mL/minute. The peak areas at these flow rates are consistent from one analysis to the next when using the same flow rate.

Any ideas or insight would be helpful. Thanks in advance.

When you increase the flow rate, the sample goes through the detector faster, so less data is collected. Since less data is collected, your response (area) is decreased.

Your detector sees concentration. Your integrator creates concentration x time as peak area. The idea of using peak area is to get mass. To get mass, you would need to multiply the peak area of your integrator with the flow rate ( mass = (mass/volume)* time * (volume/time) ).
If I do this for your three flow rates, I get 1332, 1338, and 1344, the same number within 0.5%. The remainder are either integration errors or flow rate imprecision.

Uwe - Can you show me how you calculated one of your values? I am a little uncertain what you are using for volume in your equations. Thanks in advance.

1665 x 0.8 = 1332
1338 x 1.0 = 1338
1120 x 1.2 = 1344

Thanks Merlin!

Uwe or Merlin - Thank you for your input. Can you point me to a reference from a book or publication with more information on this topic. I would like to learn more. Thanks.

Hi all, What you are describing out there is one of the most important and fundamental specifications of LC detectors termed as “detector response”. According the last definitions of IUPAC you can have:
A mass-flow sensitive detector which is “a device the response of which is proportional to the amount of sample component reaching the detector in unit time”.
A concentration-sensitive detector which is "a device the response of which is proportional to the concentration of a sample component in the eluent.
Both definitions can be found in IUPAC Compendium of Chemical Terminology, 2nd edition (1997), 1993, 65, 849.

So, for a mass sensitive detector the response is given in mv/mass/unit time while for a concentration sensitive detector the response is given in mv/mass/unit vol. According to the data that you showed it can be clearly shown that the UV is a concentration sensitive detector…
If you are looking for further details you can also see in the book: Liquid Chromatography Detectors, J. Chromatogr. Library (page 14 for detector response) by R.P.W. Scott (rather old book; 1977).

Do we agree that mass injected is constant? Let's assume that w is the flow rate. In an small time dt, a volume dV = w.dt came in the detector cell. Thus, the mass is given by dm=c.dV=c.w.dt. The problem is that c=c(t) so that c=A(t)/e.b where A(t) is the absorbance measured at the time t. Therefore dm=(w/e.b).A(t).dt. Integrating this ecuation (constant flow when substance is passing througgh the detector cell) we have the mass passing through the detector: m = w.(1/e.b).PeakArea
where PeakArea is the integral of A(t). Now we can see that peak area is inversally proportional to the flow and that´s the reason why you see that behavior. For instance, microbore columns (2.1 mm) have more sensibility in the same chrom. conditions than a 4.6 mm ID column because you have to reduce the flow-rate to 0.2 - 0.3 ml/min (so peak height increases 3 to 4 times)
Hope this helps. Sorry for my English
MFB

MFB and kostas - Thank you for your replies.

MFB - Your equations confused me a little bit, but I believe you got the fundamental answer across to me. Let me see if I understand you properly. Peak area is the integral of Absorbance over a specified time. Absorbance is obviously concentration dependent using a UV detector and concentration is simply mass divided by volume. Since the mass of each injection stayed the same (and it did in my case), concentration is inversely proportional to the volume of the analyte in the detector's flow cell. Flow rate directly effects the volume of analyte in the detector's flow cell for a given amount of time. Consequently, flow rate is also inversely proportional to peak area which explains my observations.

Is my analysis of your message correct? Please let me know.

That is right J.P., If equations is what you want I'm just copying for you the equations of the book reference that I'm mentioning above.

Concentration sensitive detector
Rc=hwQ/sm,
Mass sensitive detector
Rm=hw/sm
where h is the peak height in mV,
w is the width at 0.607 of the peak height in cm
m is the mass of solute injected onto the column
s is the chart speed in cm/min!!!(I told you the book is dated...)
Q is the flow rate of the mobile phase in ml/min

The majority of LC detectors are concentration sensitive but there are some mass sensitive detectors out there. Furthermore the test that we employing in general for characterisation of a detector is vary the flow rate and observe the peak area/height. The chemiluminescence nitrogen detector is for example a mass sensitive detector (one of the reasons it is equimolar).

J.P. I know that equations are not "very clear" but you've got the idea, as indicated by Kostas.
I think that sometimes it is better to get an idea through mathematics, although the first time seems to be more difficult. Well I better go to study English!!

MFB

As suggested by Costas, I saw a book (not by Scott, but by littlewood,1970). The difference is clear, but I am not able to find any example. WHY one detector should see mass and another one concentration, as concentration is simply mass divided by the volume of the detector chamber?

As suggested by Costas, I saw a book (not by Scott, but by Littlewood,1970). The difference between the two modes is clear, but I am not able to find any example. WHY one detector should see mass and another one concentration, as concentration is simply mass divided by the volume of the detector chamber?

If it will help clarify the issue, imagine what would happen in a detector cell if the flow were to stop abruptly.

In a mass-sensitive detector, the signal would drop back to baseline (e.g., in the GC FID detector, once you've burned all the analyte molecules, there's no more signal).

In a concentration sensitive detector, the signal would stay high (e.g., in a UV detector, the analyte is still in the flow cell, still absorbing light).

In general, mass-sensitive detectors tend to be "destructive" (FID, electrochemical, MS), while concentration-sensitive detectors are non-destructive (UV absorbance, fluorescence, RI).

-- Tom Jupille / LC Resources