HPLC instrumentation questions

[Nish]

Hi All,
I am new to this group. This is one of the best scientific discussion groups I have come across and I spent almost a day going through the past posts, and they are all really informative.

I am trying to buy a new system for our lab and was comparing the leading brands out there. I have 3 specific questions more dealing with HPLC instrumentation.

1) The representatives say that with the advanced technology now the PDA detectors have the same sensitivity as a normal uv-vis. My application, primarily with plant secondary compounds, needs a low uv detection ~220 nm. I would greatly appreciate if anyone could comment on the sensitivity issue.

2) Some of the PDA has 512 diodes and some have 1024 diodes. To me, it appears that the higher the diode elements, the greater would be the spectral resolution. But the company representatives have a different opinion. They say an element resolution of 512 diode with a 1.2 nm / element resolution and wavelength accuracy of 1 nm is good as resolution with a 1024 diode PDA. Is this right; more over does this difference in resolution make any difference in practical sense.

3) what is the main difference between a high pressure mixing HPLC system vs a low pressure mixing system. Because of the complexity of my sample matrix I usually end up with a gradient elution for analysis. Which system would be the best suited for my analysis?

I greatly appreciate any comments and suggestions.
Thank you

[danko]

Hi Nish

Here are some/my thoughts about these matters:

1. I’m almost convinced that it’s quite true. But you should also keep in mind, that other factors could affect the sensitivity as well (e.g. the band width one chooses and the flow cell volume, or rather the light path length). So you’ll need to choose the “rightâ€

[tom jupille]

Re comparative sensitivity: I haven't looked at specifications in the past year or so, but my impression is that variable wavelength detectors on the whole still have the sensitivity edge over photodiode array detectors. As always, you need to check the specs on the specific systems you are considering. Another factor to consider, however, is that if you have a multi-component chromatogram, a PDA lets you quantitate each peak at it's optimum wavelength, whereas a variable wavelength forces you to settle for a compromise, which may negate the sensitivity advantage.

Re number of photodiodes: I would assume that these are "off-the-shelf" parts rather than being custom-made for HPLC detectors. More photodiodes probably mean better wavelength resolution, but poorer sensitivity (smaller area, less light collected, worse signal/noise) and slower scanning. On the other hand, UV spectra in solution are notoriously information-poor, so that spectral resolution is less important than signal/noise (in fact, the signal from adjacent photodiodes is often combined to give better sensitivity at the price of a broader bandpass). My feeling is that lots of optical design details have more influence than the number of diodes.

Re pumping systems, again, "the devil is in the details" (I've seen good and bad designs of both types!). My feeling is that, on the whole, high-pressure mixing systems are marginally better for binary mixtures, while low-pressure mixing systems are almost a requirement of ternary or quaternary mixtures. As with detectors, general statements may not apply to the specific systems you are considering, so you need to look at the specs for things like flow and proportioning accuracy.

[Hfranz]

By design, variable wavelength detectors show less noise and have a wider linear range, so you can better detect small and big peaks at the same time.
My recommendation is: Pick a variable wavelength detector for routine analysis, a DAD for method development work. Check for good noise and linearity specs.
It is easy with the pumps:
Low-pressure gradient (LPG) pumps
- Typically 4 channels for high flexibility
- Cheaper than high-pressure gradient (HPG) pumps
- Better mixing than HPGs
- Typically optimal for use at 1-3 mL/min.
- LGPs are typically not good at very low and very high flow rates
- High gradient delay volume, typically best with conventional column IDs such as 4.6 mm
- There are pumps with and those without pulse dampers. Pulse dampers change the gradient delay volume with changing system pressure.
- Less wear parts than HPG.

HPGs:
- If you are interested in fast separations, consider HPGs with low gradient volume. Low could mean less than 100 µL.
- Good for small bore columns (e.g. 2.1 mm ID)
- Very precise even at low flows.
- Typically only two channels.
- Can typically handle higher flow rates better, up to 10 mL/min.
- Less problems with air bubbles.
- For best gradient performance, the pump should automatically compensate for changing gradient compositions.

[unmgvar]

for 1)). we actually did a study between a PDA and a variable wavelength detector of the same brand of their last generation and we saw that due to the difference in noise S/N was 6-7 times better with a variable wavelength detector. linearity of a variable wavelength detector is 2.5 AU while the PDA is at top 1.8 and best used at 1.5 AU. but in general what your rep says is correct. PDA have become better and in many cases you can use them because they do achieve the sensitivity that you required of variable wavelength detector of a generation ago.

for 2)). from my experience have never needed in truth better then 2 nm spectral resolution. i have checked several times and going to 1 nm only gave me bigger files. check for other specs of your PDA. what is the slit width? can it be changed? do you have a ref wavelenght for the 3D information? if yes then can you choose the wavelenght? can you take it out if necessary? how can you set the spectral resolution? how friendly is your software to use the information? ask them to show you how to check peak purity or to do a library search. how is a max-plot done.

for 3)). Hfranz pretty much covered it. but i'd check 2 more things:
- HPG pumps do only binary mixing but some vendors do sell a 4 channel pump which can switch in order to wash the system especially good if you are using buffers.
-220 nm sounds like you might be doing proteins or polypeptides (just a hunch). anyway i saw that for both type of pumps when using TFA if your dwell volume is actually too small then it is not good. you might need to get a bigger mixer size.
also in general HPG pump are far better then LPG for very slow changing gradients. in LPG those gradients look like steps.