Sensitivity of Peaks

[mobunvar]

@Peter

is it necessary to do with 3 micro m column. Agilent people didn't suggest me like that.

[bisnettrj2]

From your configuration, I'm guessing your system max pressure is 600 bar, right? If so, you don't have a UHPLC (typically capable of 1000-1200 bar, optimized to have low system volumes, etc). You have what Agilent used to call a Rapid Resolution HPLC. You can use their RRHT columns (Rapid Resolution High Throughput) with your system, or their Poroshell 120 columns (or similar). This system was Agilent's first attempt at cracking into the UHPLC world dominated by Waters at the time. UHPLC is basically used to describe systems that can realize the efficiencies provided by sub-2-micron particle size columns (and also superficially porous columns like Phenomenex's Kinetex, AMT's HALO, and Agilent's Poroshell columns).

But I digress. Like Tom said (and like I said previously, as well as Alexandre) - you probably aren't going to get to your desired detection limit using UV detection, regardless of the improvements you make. To get as close as possible, you can do the following:

1. Use the most UV-transparent mobile phase you can - something like acetonitrile-water.
2. Use as small a particle-sized column as you can (the 1.8-micron RRHT columns can be used on your system - something like a ZORBAX SB-C18 RRHT might work).
3. Ask Agilent about a low-volume flow cell for the DAD that will help you maximize UV-sensitivity (I'd guess they'll suggest something in the micro to nano range).

Micro: 1.7ul, 6mm path length, 400bar P/N G1315-60024
Semi-Nano: 500nl, 10mm path length, 50bar P/N G1315-68724
Nano: 80nl, 6mm path length, 50bar P/N G1315-68716

4. Make sure you are monitoring the best wavelength for each compound. You can monitor up to eight signals in the DAD - use them!
5. Inject as much sample as you can to maximize your signal-to-noise ratio, without destroying your chromatography.
6. Reduce your extra-column volume as much as possible to reduce peak dispersion.

All that being said, if you get 10 times more sensitivity than you have now, I'd be impressed. But 100? Highly doubtful. For a more in-depth look at optimizing your system, see the following:

http://www.chem.agilent.com/Library/use ... _ebook.pdf

[Peter Skelton]

@Peter

is it necessary to do with 3 micro m column. Agilent people didn't suggest me like that.

Well you're prediciting a favourable retention time window based on your technique, you said in an earlier post that 12m was too long an RT for UPLC. However, as you're not doing UPLC this should affect what you consider reasonable chromatography.

[mobunvar]

@ bisnettrj2, thank you for ur reply

My machine maximum pressure is only 400bar (not 1200 bar) and i am using Agilent 1100 series. So they prescribed me the column which i am using right now.

Any how i am planning to look for gradient elution to increase the sensitivity of the peaks.

any changes in Instrument will not possible for me because i am a PhD student and Institute is not likely to spend money on machinery. So i need to use the available condition to do better.

I commonly hear a sentence here "Blaming machine is very easy". I have to do something to increase sensitivity. what is that something, Im confused .

with the same machine, with the same column and with the same UV flow cell {I can't change any of them}, i need to develop method. I need to change my parameters for best. So shifting for Gradient Elution. Surprisingly sensitivity is increased but still gradient programe is not developed. Peaks are not separating properly.

I am unable to upload the peak picture here, otherwise u would have a clear idea about my probems.

[bisnettrj2]

mobunvar:

First of all, you're not using a UHPLC, you're using a regular old HPLC, an instrument that has been on the market for about a decade. Heck, I have a pump in my lab from 1991 that can do 420 bar. If you can scrape up $700 US (give or take, depending on who you ask), you could get a 3-micron 3.0x150mm column, but this isn't going to do wonders for your analysis, so you might as well stick with what you have.

This question is begging to be asked: Do you or your supervisor have a paper or application note stating that the detection limits your supervisor wants have been achieved on a system such as yours? If not, what is his basis for demanding that you come up with the solution?

On a general note: Try a gradient program such as the following:

1 mL/min, thermostat your column at at least 30 degrees Celsius, 10uL injection.
Hold 95:5 A:B for 5 minutes after injection
Linear gradient to 5:95 A:B over 45 minutes.
Hold at 5:95 for 5 minutes
Return to 95:5 in 1 minute
Hold 95:5 for 15 minutes

Total runtime = 71 minutes

This is obviously for scouting purposes; if everything elutes by 20 minutes, there's no need to keep the gradient going; conversely, if nothing elutes until 30 minutes, you can start maybe at 80:20 A:B instead of 95:5 A:B.

DO make sure you are injecting in a much weaker solvent than the initial conditions (or at least equal to the initial conditions) at the start of your run. If that isn't possible for solubility concerns, then you need to inject a small amount first, and then successively larger amounts until you notice a definite degradation in your chromatography.

[lmh]

This thread is rather going in circles without any great changes. Unanimously the message is that improving chromatography and experimenting with solvents and columns is not going to get you a 100-fold improvement in sensitivity.

Alexandre suggested derivatisation: if you can add a good chromophore to your target compounds, you will improve sensitivity. If you are having to work at 210nm it rather suggests your targets currently don't have good chromophores.
Several people have suggested pre-concentrating your sample so you can get a greater mass of analyte injected on column. Have you tried this?

If you cannot do any of these things, it's time to show your boss this thread, and explain that in the absence of any evidence that what he's asking you to do is possible, it's probably not possible (bisnettrj2 et al.). In the words of Scotty from Star-trek: "Ye canna change the laws of physics!". Good luck though, it's horrible being asked to do the impossible, especially in an environment that won't accept that a straightforward hplc has its limitations.