DAD detector brand influence on sensitivity

[saioa]

Dear all,

Recently I have tried to set an old HPLC method for the determination of Ibuprofen in a new system. My surprise comes when I check the instrumental detection limit. It is about 10 times higher than before!

I have played with DAD settings of the new system to improve the detection limit, with little success.

In my old lab I work with an Agilent 1100 series system with a UV-DAD 1100 detector. The column was a Zorbax eclipse DB-C8 4,6 x 150 mm column, particle size 5 um.
Now I work with a new Jasco system with MD-2010 Diode array detector. The column is a Kinetex C18 4,6 x 100 mm , particle size 2,6 um.

The method I use is the same is the same except the flow rate. For Zorbax column the flow is 1 mL/min while for Kinetex column 2 mL/min. Isocratic conditions 50:50 ACN:Buffer pH=3. Injection volume 20 uL, column temperature 40 degrees.

While in the old method (Agilent system with Zorbax column) I get an instrumental detection limit (IDL) of 22 ng/mL with the new system I obtain an IDL of 320 ng/mL. IDL was calculated as the concentration giving 3 times the baseline noise.

My first reaction to this difference was to prepare the standard solutions again from the beginning, but I get the same results. I should also mention that other compounds in the method show good IDL between 10-20 ng/mL, except for clofibric acid which has an IDL of 60 ng/mL (another method with agilent system game me an IDL of 35 ng/mL)

So my question is, if an instrument brand could give such big differences in IDL. If not…any idea why I get this big difference?

Thanks!

[DR]

I suspect it has more to do w/ age than brand. In the past few years it seems that DADs have pretty well closed most of the sensitivity gap between them and their UV counterparts.

[saioa]

Hi DR!
Thanks for the tip.
Not sure what you mean with w/ age. Do you mean the lamp age? Because it is new. We just bought the instrument and it has been running less than a month!

[DR]

I mean the difference between the new instrument and the (older) 1100...

[HW Mueller]

Dr, apparently he got the lousy results with the new instrument.

saioa, first of all you already got a factor of 2 just due to flow rate differences, you don´t tell what the first mobile phase was and whether there is a difference in absorbance of the analyte in the two liquids. You also say nothing about resolution number of plates, etc. In other words, I don´t see any basis for comparison.

[tom jupille]

First of all, look at the manufacturer's specifications for baseline noise on both systems.

Then, on the current system, get hold of the manufacturer's suggested performance qualification test for baseline noise and run that to confirm that the detector is, in fact, running within specification.

Those tests are usually specified under static (no-flow) conditions. Run the same test again with flowing solvent to see if there is a flow problem.

Now comes the hard part. Hopefully you have a record of the *exact* settings you used in the old place. These would include the wavelength, the bandpass setting, was a reference wavelength used (if so, what wavelength), and was any filtering (e.g., time constant) enabled. Go into the software on the new system and make sure that the settings are equivalent (or optimized for that detector). All of those things can have a major effect on noise.

If the specs are comparable (and your system is meeting the specs!), and you have optimized the settings, then what remains is the chromatography: flow consistence, degassing, solvent purity, etc.

[Uwe Neue]

Hmmh... One possibility is that you are at the edge of the spectrum for ibuprofen, while this is not true for the other analytes. Then the bandwidth of the DAD can play a role. Should be easy to check if this idea has merit.

[HPLCCONSULT]

"saioa": As consultants, we see this all of the time. A few comments

(1) You are comparing results using two different HPLC Methods (two different columns used). For a better comparison try using the same method/column on each instrument and see what result you get.

(2) You do not specify the DAD wavelength parameters used in either method. What is a analysis wavelength, bandwidth, is a reference wavelength used... ? These settings can have a major impact on sensitivity.

(3) You do not mention what type of flow cell is used in either instrument. Are the path lengths the same ? Are the volumes the same ? Longer path length can equal higher concentration.

(4) What about data acq settings used in the two instruments ? Changes to these can also change the outcome.

(5) Other items which can cause differences: Slit Width, lamp(s) age and stability, stability of pump...

Just some things to consider when making a comparison.

[saioa]

Deal all,
Thanks a lot for all your comments. When I posted the message I was just wondering if the DAD conditions could give a mayor impact on sensitivity because, unfortunatelly I do not have access to my old HPLC because I work in a new place and I could not compare the 2 machines. Your comments made me think about all the details used in an HPLC method. Well, after checking all DAD parameters bandwidth, reference wavelength etc. which were the same or nearly the same as in the old one and finding similar theoretical number of plates for both columns I started to think that something was wrong.
Well, I have to admit that I made a basic mistake! The buffer composition! From my old notes I could see I used 50:50 Buffer pH 3 and Acetonitrile. So I prepared a formate buffer pH 3. Afterwards checking my old quality control documents I saw I used phosphate buffer. I tested phosphate buffer and bingo! the sensitivity for Ibuprofen increased to the same levels as before or even a bit better!
Thanks again for all your help.