Difference in Response Between Methods

[BenNC]

I have a compound with a really lousy chromaphore forcing me to monitor at 200 nm. I currently have two methods that I am using. One is for release and stability testing (AIM) and the other is for in-process control (IPC) during manufacture of the drug product.

The AIM method is optimized for separation of impurities and the IPC method is optimized for speed. The real differences between the two methods are in the organic percentage (30% for AIM and 33% for IPC) and in the flow rate (1.0 mL/min for AIM and 1.5 mL/min for IPC). Both methods are isocratic.

During analysis, I see a difference in response between the two methods when analyzing the same solutions. The response for the IPC method is about 70% that of the AIM method. I believe that I am seeing the result of a shift in the spectrum between the two solutions, but if anyone else has any ideas I would welcome the input.

Anecdotally (sorry, but I don't have a lot of data for both methods on the same lot of material), I seem to be seeing a high bias (1-2%) on the calculated results for the IPC method compared to the AIM method. I would welcome any theories as to why this may occur. To get rid of an easy answer though, the closely eluting impurity is still resolved from the main peak and it is also present in the standard, so any error from coelution would be mostly cancelled out.

Thanks in advance,

Ben

[Uwe Neue]

Here is another idea: I assume that you are comparing peak areas obtained with one method to the other method. Peak area at equal amount injected depends on the flow rate. Thus the faster method would give you 1/1.5 = 67% of the peak area of the slower method.

[BenNC]

Uwe,

I didn't realize that flow rate would have that much of an effect on response. Foolishly, I will admit to having some skepticism. Therefore I did a couple of injections to verify.

The ratio of the 1.5mL/min response to the 1.0 mL/min response was 65%.

Despite the data in front of me, this does not really make sense to me. The same amount of analyte is injected onto the column and therefore passing through the detector in both cases. What is causing the difference in response?

[Uwe Neue]

Most HPLC detectors, but especially the UV detector, have a signal that is proportional to the concentration in the mobile phase. You integrate the peak to get a signal that is proportional to the mass of analyte in the peak. However, if you increase the flow rate by a factor of 2, the signal flies by the detector 2 times faster and is therefore 2 times narrower. Thus the integration is 2 times less.

In math: you integrate AUFS x time. AUFS = absorption coefficient x pathlength x concentration. Thus after calibration, your signal is concentration x time. Concentration is mass/volume (=ml). In order to get mass, you need to muliply with flow rate, which has the dimension of mL/min.

((mass/mL) x minutes) x (mL/min) = mass

[HW Mueller]

If you had an internal standard in there it would have pointed that out for you.

[BenNC]

So is this basically an effect of the dwell time in the flow cell?

BTW - The anecdotal bias between the methods mentioned in the original post does not hold up under the scrutiny of actual numbers.

[Uwe Neue]

No, it's not the dwell time in the flow cell. It is the integration that is screwed up.

[Bill Tindall]

I would call it a dwell time effect. The signal is proportional to the amount of time the stuff dwells in the detector. The number of absorbing molecules is the same in either case but how long they spend(dwell) in the detector path is different.

The way I try to get people to understand it is to think what the signal would be if you turne dthe pump off while the peak was in the detector. If it is obviouls that turning the pump off would increase the signa, it is a small step to see that slowing the flow will do the same thing.

[Bill Tindall]

I would call it a dwell time effect. The signal is proportional to the amount of time the stuff dwells in the detector. The number of absorbing molecules is the same in either case but how long they spend(dwell) in the detector path is different.

The way I try to get people to understand it is to think what the signal would be if you turne dthe pump off while the peak was in the detector. If it is obviouls that turning the pump off would increase the signa, it is a small step to see that slowing the flow will do the same thing.