As a follow up to the discussion in comments: it's now clear that UV peaks will have larger area if flow rate is decreased, because the same portion of molecules are measured again and again.

With Mass Spec it's different because it's a destructive detector. So does the flow impact peak areas in Mass Spec? And how much does this impact differ from UV?

  • 1
    $\begingroup$ If the same molecules are recorded «again and again» (sorting the column, passing the UV-Vis detector), than why shouldn't the TIC integrated over time when these molecules pass the MS detector equally rise? $\endgroup$
    – Buttonwood
    Sep 16, 2021 at 8:29
  • $\begingroup$ @Buttonwood, since I've never actually worked with the instrument let alone looked under its hood, I may easily mistake here: I thought that the whole flow that comes out of the column is directed to MS. Then it's ionized and injected into the mass analyzer. If we decrease the flow, we don't change the number of particles passing to the analyzer. Since MS just counts the number of particles (which didn't change), we should end up with the same area. And after the particles passed the detector, they're "destroyed" and never return to the detector which means they aren't counted again. $\endgroup$ Sep 16, 2021 at 8:45
  • $\begingroup$ @Buttonwood, oh, or I may have misunderstood your question. With UV-Vis detector we have a flow cell and the detector is shining light through that cell. The longer the molecules pass through that cell the more they're exposed to the detector. With Mass Spec there's no flow cell - it just takes the particles from the chromatographic column and "counts" them. This process means that the particles fall onto to the mass detector (or are filtered out) and get out of the system. So they have only one shot at being measured. $\endgroup$ Sep 16, 2021 at 9:08
  • $\begingroup$ The flow rate of the mobile phase (like a carrier) affects when the molecules sort the column, broad or narrow peak shaped, and (van Deemter equation) the quality of separation of a mixture passing the column. UV, MS, FID; these all are merely observers at the exit gate (cf. e.g., HPLC simulation on ChemCompute). Different to UV and FID, for HPLC-MS, only a tiny split portion sorting the column actually enters the MS detector. In total (i.e., over time), the same number of molecules pass column and (fixed portion) detector, high, or low flow rate. $\endgroup$
    – Buttonwood
    Sep 16, 2021 at 9:13
  • $\begingroup$ If the molecules of interest sort the column like a flat broad peak, then the recorded signal for a brief increment of time (say $\pu{10 s}$) might be well different to an increment of recording the TIC for a $\pu{10 s}$ interval about a narrow, sharp and high band of molecules sorting all at once the column. And true, once hitting the sensor, the molecules are «dead», and not accumulated. The Faraday beaker isn't about collecting the molecules to fill a vial. $\endgroup$
    – Buttonwood
    Sep 16, 2021 at 9:18

1 Answer 1


As we discussed previously in the comments given in your link, in the analytical chemistry literature, you will see two types of detector classifications especially in the field of chromatography, namely, concentration sensitive and mass flow rate sensitive detectors. This terminology was introduced by Istvan Halasz in the 1960s.

When we wish to classify detectors as concentration sensitive vs. mass flow rate sensitive detectors, we have to eliminate the column discussion in between. This is the stumbling point and therfore the concept is taught or discussed quite incorrectly even by highly experienced chromatographers. So how are we supposed to call HPLC-UV detector as concentration sensitive vs. HPLC-MS as mass flow sensitive detector?

Do a thought experiment. Fill a HPLC eluent bottle with a solvent and add a few crystals of benzoic acid and dissolve. Connect the pump to the detector directly and monitor the absorbance at 254 nm as a function of flow rate.

At 0.05 mL/min the absorbance is $A$, 0.25 mL/min absorbance is still $A$, and 5 mL/min, it is still $A$. This is because Beer's law does not have a flow rate term in it. Now dilute that solvent containing benzoic acid by half and repeat the same experiment. The signal is now $SA/2$, and same observation is to be made at high flow rates.

Thus, HPLC-UV detector is a concentration sensitive detector.

Come to a GC-MS system (I am avoiding HPLC-MS, see below). Use an empty silica column as before. Inject the mobile phase containing a small amount of benzoic acid as before. At 0.05 mL/min, the TIC signal is $S_1$, increase the flow rate to 0.15 mL/min, the signal is $S_2$. It turns out that $S_2$> $S_1$. Mass spec it therefore mass flow rate sensitive detector, because by increasing the flow rate more molecules are reaching that detector and the response will increase. In this classification, the column has no role and we are only interested in the signal magnitude not the peak areas , because there is no column.

The story changes, when you are interested in peaks areas and when there is a column. This classification, like every human classification, breaks down.

Let us reconnect the column and use HPLC-UV detector. Use a clean solvent, and inject a solution of benzoic acid. You should see one peak with area $X_1$ at 0.25 mL/min, and for the same injection, at 1 mL/min, the peak area would be $X_2$. It will be found that the peak areas $X_1$>$X_2$, the reason is not that the same portion of the analytes trapped in the flow cell are measured again and again, the reason is that the residence time of the analytes is larger at low flow rates. Peak area is the integral of the Signal x Time! More time means larger peak area.

The story with mass spectrometer is slightly challenging with flow rates and much less predictable, because now we have two coupled effects

(i) Mass flow rate effects, that the signal should increase with flow rate because more molecules are reaching.

(ii) Peak area should not be affected too much from flow rates because the residence time of the analyte in the mass spectrometer is too short regardless of flow rate. It is a destructive detector unlike the UV. So residence time does not matter much,

So, it is hard to predict what will happen to the peak areas in MS because a third complicating factor is the ionization efficiency may change as a function of flow rate.

To show you what a mess HPLC-MS is, look at the following graph. The peak area is non-linear function for a very common HPLC mass spectrometer, the ESI.

enter image description here

Figure taken from Combined Electrospray Ionization−Atmospheric Pressure Chemical Ionization Source for Use in High-Throughput LC−MS Applications; Richard T. Gallagher et al. in Anal. Chem. 2003, 75, 4, 973–977 (doi 10.1021/ac0205457).

In short, Halasz was an excellent separation scientist but his classification breaks down in real life, yet he did a lot of good work. I heard that he commanded good respect in his circles in the 1960-70s.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.