Running 3 mm ID columns on Standard LC Systems

[lmh]

Danko, entirely agreed! Where I work, no one wanted to be tied to a single manufacturer for all their columns, so we never bought UPLC. Curiously, if Waters had engaged with other column manufacturers to ensure alternative supplies of sub 3u particle columns, we might well have bought their instrument.

[danko]

Hi Anthony_Ng,

The main problem with your investigation – as I see it - is the column length.
Try a 100 mm version of the 2.1 mm diameter column and you’ll get the needed resolution.

Best Regards

[Uwe Neue]

LMH: here is a list of the UPLC columns available from Waters:
ACQUITY UPLC® BEH C18
ACQUITY UPLC® BEH C8
ACQUITY UPLC® BEH Phenyl
ACQUITY UPLC® BEH Shield RP18
ACQUITY UPLC® BEH HILIC
ACQUITY UPLC® BEH Amide
ACQUITY UPLC® BEH Glycan
ACQUITY UPLC® BEH130 C18
ACQUITY UPLC® BEH300 C4
ACQUITY UPLC® BEH300 C18
ACQUITY UPLC® HSS C18
ACQUITY UPLC® HSS T3
ACQUITY UPLC® HSS C18 SB

I think this is a reasonable selection. I am looking forward to your suggestions on what we need to add...
Also, you can use any column on a UPLC system. You may loose the capability to run at UPLC pressures, if the column is not rated for this, but nobody hinders you from connecting it.

[unmgvar]

Anthony_Ng

you are shpowing us chromatograms from 2 different detector correct?

the first one from a VWD-monochromator
the second one from a DAD.

your first main issue is that of RI difference due to optics differences between VWD and DAD.
it shows with your solvent peak. this is very related to another posting we have been discussing called "agilent detectors" currently in page 2.
i am still not convinced from it that you can safely swap between types pf optics has agilent like very much to promote. we have our shares of problems with such an issue currently.

the rest of it does shows that you probably very well have some dwell volume somewhere, most probably your flow cell. it is too big.
you should go for a semi micro size of around 2,5 ul and not a semi analytical of 5 ul.

[unmgvar]

lmh,

i agree with you on the need to be less dependant of one vendor yet these days there is enough column and UHPLC hardware competition out there.

my main problem with UHPLC's are the salesmans and marketing info on these products.
you should very carefully weight what they tell you. from my experience they do not yet have really enough knowledge of what it is and what it can do in reallity and what are the different needs of such systems. sometimes they trully tel you real what they want and it completely oposite to reality.
like anthony for example who most is probably using a not adequate enough flow cell size for what he is doing.

when you know the systems capability and limitiation it can be a very nice addition to your lab if there is such a need for it.

[Kostas Petritis]

Where I work, no one wanted to be tied to a single manufacturer for all their columns, so we never bought UPLC. Curiously, if Waters had engaged with other column manufacturers to ensure alternative supplies of sub 3u particle columns, we might well have bought their instrument.

UPLC is here to stay and I anticipate the whole industry to shift eventually into it. Several manufacturers already offer such instruments and consumables (columns) and I would anticipate more of them to come up with relevant products this year...

[Bryan Evans]

Anthony -

Just curious, have had any success yet? If you're still stuck,
75x3mm (3um) is another option. The larger dimension may
be more forgiving.

[Anthony_Ng]

unmgvar,

I use DAD for all runs. But follow Agilent application note, switching DAD data rate faster (they have 20Hz by default, I change them to smaller peak width, 40Hz)

I guess there maybe some extra column volume somewhere. I didn't change the small tubing from needle seat to Rheodyne injection port yet. Also flow cell is another issue I think.

Bryan_Evans,

I haven't got success yet. The story is.....this instrument is LC-MS. Sales person suggest 150mmx4.6mm 5um column and optimum flow rate is 0.4ml/min for ESI/APCI ionization.

Our professor thought it is too slow to run a sample (around 30min). So I bought a 50mmx2.1mm 1.8um, hoping that we can shorten the run time but keeping resolution acceptable.

I am thinking about switching to 75mmx3mm 3um but professor will surely blame me waste of money.

Anthony

[Anthony_Ng]

Here is a better separation of the short RRHT column.

But condition needed to be changed to: 57:43 MeOH:H2O, flow rate 0.3ml/min

It looks a better separte for early peaks but for Benzonphenone and Toluene (RT ~2.5min) started to merge.



Just cannot achieve the manufacturer claimed "Plug and Play" column change without adjusting method condition.

[unmgvar]

Anthony i was refering to the chromatogram done with the analitycal column. it is done on a HPLC with a normal UV or a DAD?

in 99% of the cases that are like yours with these amounts of peaks "plug and play" is non existent.

for one peak plug and play might happen,
for a few peaks maybe
in any cases most of the time you will always need to adjust method transition from the various tools provided by the vendors.
all cases that i have seen presented were always easy to do cases with sufficient resolution or other requirements so that transfer was easy.
the real world is where you will need to optimise your method almost always, and that's because most chemistries of columns are not entirely equivalents even from the same vendor. there are differences almost always in particle distribution, carbon loading, special silica treatments added and hardware have different pumps with new dwell volumes

[Uwe Neue]

In an isocratic chromatogram like this with similar molecular weights you can calculate the plate counts for the peaks from beginning to end. You will get an improvement. This is due to extra-column bandspreading on the early peaks. You can then even make an estimate from this information about the extra-column bandspreading in the instrument by plotting the variances against the retention times squared. There is some scatter due to differences in the diffusion coefficient of the different analytes, but with some judgement, you can get a reasonable value of extra-column bandspreading.

[Anthony_Ng]

I looked again the column performance report (shipped with the column) that:

Both sample components are the same: Uracil, Phenol, 4-Chloronitrobenzene, Toluene

For 4.6x150, 5um column:
Mobile phase: 85:15 MeOH:H2O
Theoretical Plates: 15958
Selectivity: 1.75
k': 1.07

For 2.1x50mm, 1.8um column:
Mobile phase: 55:45 ACN:H2O
Theoretical Plate: 11104
Selectivity: 2.05
k': 6.70

Just looking at theoretical plate, can I say that the shorter RRHT column has worse efficiency than the long column, leading to the worse resolution of the previously posted chromatogram?

But the mobile phase are totally different. Can we just compare the Theoretical plate and come up with the clue?

Thanks again for answering those simple questions because I was in organic synthesis background. Just have 1 year HPLC experience.

[Uwe Neue]

Looking at your chromatogram, I am still convinced that you still have too much extra-column bandspreading. As mentioned above, you can either measure it or you can try to get at it from a recorded chromatogram. In order to get it from the chromatogram, do a plot of retention volume squared divided by plate count versus retention volume squared. Draw a line through it. You will get a positive Y intercept, and this intercept is the squared value of your system bandspreading (in volume units ^2).

[Anthony_Ng]

Thanks Dr. Neue

I am not quite sure myself. I used 5-Sigma method (from Waters "care and use" manuals) to calculate System Bandspreading Volume:

Sample: Acetone
Solvent and mobile phase: MeOH
Flow rate: 0.4ml/min

Hence I followed the steps and find that Bandspreading Volume is around 35ul.

[Uwe Neue]

If I understand you correctly, this is about as good as one can get using this instrument. However, your early peaks are dominated by extra-column bandspreading. With the instrument that you have, you won't be able to go much further.