Chromatography Forum

HI folks..

Here is the issue I have...

I did a drift and noise test on my HPLC w/ 1047A RI detector and here are the results I got :

Noise : 5.148e-2 mAU
Wander : 6.646e-2 mAU
Drift : 1.948 mAU/h

Now the specs for the RI in the manual are :

Noise : 5e-9 RIU/P-P
Wander : 2e-8 RIU
Drift : 2.5e-7 RIU

How can I convert these??

My RI detector states that integrator is calibrated at 1x10-5 RIU = 32 mV

So assuming that 1 mV is 1 mAU (advice from another member)

NOISE => 0.1933 mAU = 6.64e-4 RIU

DRIFT => 1.948 mAU/hr = 6.70e-3 RIU/hr

Wander => 0.0665 mAU = 2.28e-4 RIU

Does this seem right or am I missing something?

Please let me know if anyone has any information that can help me out..

Thanks
- Hubert

I read your other post and maybe I'm missing something...why are you getting a signal in mAU for an RI detector? If you're using an integrator or A/D convertor for the 1047 shouldn't you be getting a signal in mV?

Also, it's not always 1 AU = 1 V. We have an Agilent 1100 multi-wavelength detector, and the settings for the Analog output are 2.5 AU = 1 V or 0.1 V, whichever you select.

Hi Noser,

Thanks for the info.. I did some more digging into it and found a couple things that seem strange to me..

Here is a capture of the RI properties as well as a baseline, pump pressure and flow :



It looks like the flow is OK, but the pressure is up and down, now this is an isocratic pump, but I don't think I should see variations that look like an oscilating line.. am I right here?

In regards to the RI settings, the setup shows 2000 U/volt (I'm assuming these are the AU's), does this make sense?

Thanks again for all your help..

- Hubert

The U for your detector is Refractive Index Units so the output is 2000 RIU/V.

As for the ripple, that can be caused by many things, but the top of the list include:
1. flow ripple from a bad check valve or bad seal
2. thermal mismatch between the mobile phase and the detector
3. a tiny bubble stuck to the flow cell
4. excess backpressure on the flow cell

hmmm... that indeed is very interesting..

I don't believe seeing this before, but this is a neew phenomenon since I started investigating this... would this have an effect on my peak areas? Latelly, I can't get the reproducibility on the area's under the peaks and I was thinking that this could be due to the drift and noise on my detector.. what should I invesigate first??

Thanks
- Hubert

Yes, severe ripple will cause the baseline to be drawn at odd places. Is the percentage of ripple versus peak height about the same magnitude as your area variability? If so, fix the ripple first.

so... what's the easiest way to investigate this??

How can I determine what the issue with the pressure is? How can I check if the check valve is bad?

Any other suggestions I can look at?

- Hubert

Those settings in your ADC are customizable...looks like either those are the default or someone had a UV detector in there before.

You could change the "Unit" to RIU, and change the "Units/Volt" using the conversion from the manual 1x10-5 RIU = 32 mV.
Of course, if you routinely attach different detectors to the ADC then other users need to be aware of this and incorporate it into the method files, or just leave it the same and manually convert whenver you need to know RIU units or mAU.

RI detectors are very sensitive to pressure fluctuations, so I'd check the pump first as Mark said. If your Agilent came with a routine maintenance video, it shows you how to do this.

nope.. no videos this is an old unit we bought 9 yrs ago.. it's an 1100 series HPLC with a 1047 A detector (don't ask me why..) my other HPLC is an 1100 series complete, and so fra that one has been good to me..

Anyone know where I can get my hands on one of those videos on pressure issues?

- Hubert

Another thing, what type of column and mobile phase is that?

It looks to me like you're doing carbohydrate analysis with a ligand-exchange or ion-exlusion column.

87 bar = 1262 psi

That's pretty high for these columns. Do you have a guard on it? Maybe it is time to clean or change the guard and possibly the column.

We have a guard column on the main column. The pressure is usually in the 77 bar range (it's kind of high now as well.. so there could be a plugginng issue after all).

Both the columns are almost brand new.. about a month old or so.. we usually get 1-1.5 yrs out of these babies..

- Hubert

I haven't used an 1100 for years, so I forget the details. There is a diagnostic feature called ripple. Find it. The spec is around 2%, but I know that a pump in good tune can do less than 0.5%. (The old 1050 could do that.) These pumps have a semi-automatic diagnostic procedure to test the seals and check valves; that is documented in the manual. If you don't have one, call your Agilent rep. While you are checking things, there is a filter in the pump head; see if it is dirty; that could account for the slightly high pressure. Also, check the detector flow cell and make sure there is no restriction in it.

well... I sonciated the ball valve and it seems like the pressure oscilations are a thing of the past.. now I have a steady pressure line.. I will let it flush overnight and test it out tomorrow. I will also do the ripple, and leak tests in the morning as well to see what we have there.

It recomends using IPA to do the pressure and leak tests, is this a must or could I use another alcohol like ethanol? If not, what is the reason?

You guys have been a great help.. thanks again for helping me out.. as you can tell I'm somehwat new to this and learning by doing it...

Thanks again..

- Hubert

The tests use the loss of pressure over time to measure the leakage; that depends on the viscosity of the fluid. So the factory spec assumes the viscosity for IPA.

Hi folks..

I finally got around to doing the leak and pressure tests.. and guess what.. I got both fails on the same item :

'Final Pressure drop with flow 0 : should be < 2bar/min --- mine was 3.8 bar/min

After a number of tries I finally got it down to 3.2 bar/min but not any lower.. any ideas what could be the cause here?? I'm assuming this is due to a leak and the pressure drops too fast, but where.. this seems like a small leak..

And the fact that this was an issue in both tests leads me to belive that this is trully an issue with this machine..

Anyone have any ideas as to where to start??

Thanks..
- Hubert