Chromatography Forum

I worked on an assay/imp method recently with following conditions:

MPA: H2O:ACN: Formic Acid=99:1:0.2 (v/v/v)
MPB: ACN:Formic Acid= 100:0.2

Wavelength: 275 nm
Sunfire C18 column

Gradient
Time A B
0 100 0
8 100 0
25 60 40
27 60 40
28 60 40
40 100 0

Different detector has different baselines. On systems (2) with VWD UV detector, the baseline started hiking at 12 minutes. It makes sense since the gradient starts and ACN percentge increases. In contrast, on systems (2) with DAD or Multiwavelength detector, the baseline started going down at appox 12 minutes. All the analyte peaks did elute at the same time. Same column and mobile phase were used for all test injections. All four Systems are Agilent. Have you seen this before? What's behind this?

Next I will try on a system with VWD and DAD (or MWD) in series to confirm if it is really due to the detector.

Duantech,

Do you use same Reference wavelength for both (VWD and DAD) detectors?

Regards

Do you use same Reference wavelength for both (VWD and DAD) detectors?

Actually, I think that most variable-wavelength detectors are true double-beam; the use of a reference wavelength in a DAD accomplishes the same thing. That said, I concur with zokitano that the changing the reference wavelength on the DAD is probably the first thing to look at.

Double beam configuration in most HPLC detectors does not mean that the reference beam compensates for the mobile phase absorbance fluctuations, but it only means that changes in lamp energy/radiation intensity would be compensated for.
The configuration is typically as follows: One of the beams hits the flow cell and the other is sent directly to the sensor - so the reference beam never transverses the mobile phase.

The other configuration is a bit different: Both beams transverse the flow cell but they are “tunedâ€

thanks all your replies.

I did not set reference wavelength for either detector. A genera question, when shall a reference wavelength be set?

All replies so far lead my thought on my another observation recently. Same solution was injected on same column on two different systems (both agilent). One is MWD, the other DAD. Instruemnt methods were set up in a same way. The response of the analyte from the DAD system almost doubled that from MWD. I know we usually don't compare reposnse between systems. Bu the difference was so huge that my curiosity could not let it go. Discussions here pointed to lamp energy difference and dirty flow cell on MWD. Could above possilities have caused such difference? Or is there anything else behind this, like a reference wavelength is needed here?

meanwhile, I will be still looking forward to your thoughts on firstl question

thanks a lot

Our agilent system with VWD (not MWD) does not have a reference wavelength seting option. After checking the manual, it seems that there is a splitting of beam at one point: one beam continues through the flow cell path, the other reflected to a sensor. Just as Danko said, the reflected beam does not go through the MP (or cell). So the correction on VWD is done by the reflected on the same wavelength through air (or void). So there is no choice of wavelength on VWD. The correction is more like for noise caused by electric variance. (?)

For MWD, after reading some text, it seems to me that continous spectrum can not be achieved unless the flow is stopped (or peak is frozen). DAD can obtain continuous spectrum with flow going. MWD can obtain up to data from five channels (our system). Only one beam goes through path for dispersive detector (MWD and DAD). Referencing is achieved by comparing the data at selected wavelength to that at a reference wavelength. So the correction is more like to take away the noise arising from the mobile phase change (mostly gradient?).

Above is just my understanding. I will really appreciate it if you can let me know about its correctness.

Back to my original question, the baseline's going down happened even no reference wavelength was set for either MWD or DAD. So the mystery still exists.

Best

Lijun

the baseline's going down happened even no reference wavelength was set for either MWD or DAD.

According to the Agilent 1200 DAD manual (page 77, Figure 30), that is exactly what you would expect to happen with no reference wavelength set:

Yes, fine example Tom.
In this particular case, the correction is made for refractive index fluctuation (at 267 nm none of the mobile phase components absorb light), since the concentration of phosphate is decreasing over the gradient.

Duantech, when you say “Same column and mobile phase were used for all test injectionsâ€

Thanks for the example from Tom. I will try with one reference wavelength.

For Danko's question, I installed the same column on the different systems with different detectors and moved around the same mobile phase at the same time.

thanks

thanks all your replies.

..... The response of the analyte from the DAD system almost doubled that from MWD. ....Bu the difference was so huge that my curiosity could not let it go. .........


thanks a lot

Just found out the cell path for MWD is 6mm and DAD 10 mm. May it have explained the difference?
best

duantech

The response of the analyte from the DAD system almost doubled that from MWD. ....

Just found out the cell path for MWD is 6mm and DAD 10 mm. May it have explained the difference?

Yes, the length of the lightpath explains the higher signal – no abrah cadabrah in this case.

Going back to your original question:

With the described mobile phases I would expect the baseline to climb up when the B eluent is increased (no reference wavelength correction).

So I thought: Could it be that the reference wavelength correction on the DAD is switched on and the degree of compensation is set too high?
I’m not too familiar with Agilent instruments so I don’t know if the degree of compensation could be varied, but this might be a viable explanation.

I sow the later post stating: the baseline's going down happened even no reference wavelength was set for either MWD or DAD.
But you might like to double check it.

Best Regards

I tried a couple injections without and with reference wavelength (380, 20). Although the baseline still went down after the gradient started, the extent of absorbance down toward negative was much less than without reference wavelength. However, the question still lingers. Why hiking baseline with VWD after gradient starts while going down baseline with MWD/DAD?

best

duantech

We have a couple of 1100 DAD detectors so I went in the lab and took a look at the optic part on one of them. While I think the flow cell is mounted in a very smart manner (easy to handle) I also think it sits a bit loosely and thus there is a possibility of small angle variation in its positioning.

So I thought: Could it be that the beam is refracted at different angles on your 2 detectors? The difference wouldn’t be observed in an isocratic mode baseline, but in a gradient mode where the mobile phase refractive index changes, it could cause alteration of the beam direction but not necessarily at the same angle.
My suggestion would be to run one or two gradients on the system with the DAD then take the flow cell out then mount it again and run the gradient again. It is very easy to detach and attach the flow cell – only takes a couple of seconds.
Then compare the data and see if the gradient profile changes – hopefully to an upwards going baseline.
It might be practical to disable the reference wavelength in this test in order to observe the real differences before and after the flow cell adjustment.

Good luck

P.S. Please get back with the outcome of the test, if you chose to conduct it.

Danko,
Unfortuately, I can not conduct the exp due to our GMP environment. Thanks for your suggestion just the same.

best

duantech

Hi

Well I would like input certain observations in our lab…during our RP-HPLC analysis using Agilent MVW detector (detection in 215nm, 210nm, 280nm with 360nm as reference wavelength) we found high noise in the analysis, when the reference wavelength was absent this noise level decreased (we got a very smooth baseline).

Could any one suggest the reason for this obervation??

And how do you set the reference wavelength nm???