trouble with RSD

lmarotta

<p><span>Hallo,</span></p><p><span>I would like to have some advice about a problem (actually it's similar to jrich's one):</span></p><p><span>in order to reduce RSD value, I have reduced the drawing rate to 10uL/min (I'm using a 1uL loop , the injection mode is PLNO and the sample volume is 0.7uL)</span></p><p><span>The RSD has improved, but it takes about 5 minutes to perform an injection (the run time is only 1 minute); </span></p><p><span>particularly, drawing post air gap takes a long time even if </span><span>I've specified a post-aspirate value of 0.0.</span></p><p><span>How can I reduce injection time and mantaining a good RSD?</span></p><p><span>These are some information about my method:</span></p><p><span>Strong wash: 400 uL (water:methanol:acetonitrile:isopropanol:25:25:25:25)</span></p><p><span>Weak wash:800uL (water:acetonitrile 75:25)</span></p><p><span>Mobile phase(isocratic) : Phosphate buffer 0.005M pH 5.5 (75%) acetonitrile (25%)</span></p><p><span>column:BEH Shield RP18 2.1 x 50 mm (40°C)</span></p><p><span>Thank you very much</span></p>

rune

Hi

A couple of questions first : Are you sample limited ? Would it be possible to change the injection volume to 1ul insted of 0.7ul ? Can you reduce the wash volumes to the default values of Strong = 200ul, weak = 600ul, or maybe even further without affecting carryover ?

Suggestion : Inject 1ul with full loop mode. The default overfill factor is 5.8 meaning that you will use 5.8 ul sample per injection. Use the default syringe draw rate. This will speed up the injection process considerably and also improve your RSD. If you need it to be even faster use the load ahead function with a loop offline time of 0.1 min.

In PLNO mode the post air gap specification is not used. Since the needle does not go into the wash station when the sample syringe moves it will automatically draw air into the needle after the sample. What you see in the console as drawing post air gap is actually the positioning of the sample in the loop, so your specification of post air gap as = 0 has no impact on the injection time.

I hope this will help.

best regards

Rune

lmarotta

Hi Rune,

Thanks a lot for your advice.I have permormed some test following your indication, and I'd like to share my results.

1) injecting 1 uL in full loop mode I obtained a very good RSD (0.12%), but unfortunately the peak shape was orrible (a 1:1 dilution didn't solve the problem)

2) injecting 0.7 uL in PLNO I had a RSD=0.15% using a 10uL/min draw rate (but it's too slow) , but increasing the draw rate to the default value I had a RSD=1.34%, and the average area decreased of about 12%.

3) reducing strong wash and weak wash volumes to the default values had no impact on RSD. I also increased the percentage of methanol (a good solvent for my sample) up to 75% in strong wash solvent, but

it did not improve RSD.

Please let me know if you have any other suggestion

Thank you very much

Lorenzo

rune

Hi Lorenzo

What did you use to dilute your sample ? The bad peak shape is that tailing or fronting or what ? Chromatograms of your different experiments would be helpful.

Best regards

Rune

lmarotta

HiRune,

I used a NaH2PO4 0.005M solution(75%)-Acetonitrile (25%) to dissolve the samples.

please find attached 2 chromatograms(1.0full loop and 0.7uL injection)

Thanks,

Lorenzo

rune

Hi again

I guess that the 1ul full loop chromatogram is of the un-diluted sample, but you say that the 1:1 diluted looks the same. To me this looks like the situation you would see if you dissolve our sample in too strong a solvent, but in this case you are not since the sample diluent is the same composition as your mobile phase.

I have no clue as to what is causing this bad peak shape, can anyone else help ?

best regards

Rune

lmarotta

Hi Rune,

I confirm the 1uL chromatogram is the un-diluted sample. Anyway, If the bad peak shape problem can't be solved, I would go on with 0.7uL PLNO injection and try to optimize the flow rate and/or SNW-WW composition and volume.

Thanks a lot,

Lorenzo

lizh

One of our checmists suggested that I ask you for the calibrated volume of the 1uL loop. She was wondering if the loop volume is small and therefore the 0.7uL volume is too high for that loop. Has he tried a full loop injection? Has he tried a 2uL loop? Also, what are his % RSD values (both before and after he changed the draw rate)?

Many TX

Liz

lmarotta

Hi,

the calibrated volume of the 1uL loop is 1.5uL. I've tried a full loop injection, but the peak shape was bad (you can see the attached chromatogram in my previous reply). I've tried a 2 uL loop, but I had RSD=1.5%

Using a 10 uL/min flow rate I had a RSD=0.15%, while with the default draw rate value the RSD was 1.34%.

Thanx,

Lorenzo

lizh

Hello

Apologies for delay - I ahve neen travelling these last few weeks

Can you confirm you injection mode PLNO or PLPA and the wash solvents and volumes.

Liz

lmarotta

Hi Liz,

My injection mode is PLNO and the SW(400uL) is H2O:Methanol:acetonitrile:Isopropanol (25:25:25:25) (I''ve tried to increase the % of methanol-a very good solvent for my sample- till 75% but the RSD was the same)

The Weak Wash solvent is H2O: Acetonitrile 70:30 (800uL)

I've also movified the WW and SW volumes to default values, but it had no impact on RSD.

Many thanx

Lorenzo

lizh

Hello again

Firstly, apologies I keep asking the same questions - I forgot to review the earlier posts. Tank you for being so patient.

Now I am back in the office I have taken a small survey of our chemists and we think it is simply an overload issue. Can you dilute your sample with buffer 50% and see what the peak shape looks like with full loop mode.

The chemists feel your other settings are fine for washes etc. But as you still have an odd shape - we are suspecting overload. The area of the 0.7 uL is 81K and the full loop 200K area counts.

Next if that works go back to PLNO after that. One thing at a time. If that is this is the issue we can continue with the dilutions and use a larger loop (5 uL say) and then we can go to a faster draw rate and get the cycle time back. As those small loops can be slow - especially at 10 uL/min.

Best regards,

Liz

lmarotta

Hi Liz,

I've just tried to dilute my sample 1:1 and perform a full loop injection.Unfortunately the peak shape is still very bad: do you think I should go on with further dilutions? (consider at moment the sample concentration is already quite low, 0.06 mg /ml)

Please let me know,

Tx,

Lorenzo

lizh

Yes please try even lower.

Liz

Newman768

Hello Imarotta,

have you tried to dilute your sample 1:1 in a way that you have 80% of water within the sample solution and inject in full loop mode?

Your sample solution (as far as I remembered) contains NaH2PO4, which has a totally different pH value compared to your mobile phase (pH 5.5). If the pH is important for your compound, than it is a basic compound and the lower pH value of sample solution leeds to the result that part of compound eluates a little earlier than the main of compound.

This could be solved by using same pH for sample solution and mobile phase OR by making the sample solution less retentive by using higher amount of water in sample solution compared to mobile phase.

Andreas

lmarotta

Hello Andreas,

I've performed some tests following your indications, and here's the results.

Changing sample solution's pH has no impact on peak shape, while dissolving the sample in water-80% / Acetonitrile-20% the peak shape is perfect (full loop mode), and the RSD is 0.14%.

Besides, I mantained the original concentration, so probably it wasn't an overload matter. Infact, diluting the sample with the original sample solution, the peak shape remains very bad, even at very low concentrations (I' ve performed tests till 0.03 mg/ml)

Many many thanks for your precious advice,

Lorenzo

Newman768

Hello Lorenzo,

it's good to read that it works with 80% water.

Does that mean you're using full loop mode with normal sample drawing speed? Or are you using the slowed drawing speed?

Andreas

lmarotta

Hi Andreas,

I'm using a normal draw rate. It allows me to have very fast analysis, considering my run time is about 1 minute.

Thank you again,

Lorenzo

ElseKragt

Maybe you can use the Partial Loop mode. Then the loop will be filled with sample and weak wash solvent, instead of sample and mobile phase as it is using the PLUNO mode.