Chromatography Forum: LC Message Board: Low level anion analysis
------------------------------------------------------------------------------------------------------------

By Charles Wilson on Thursday, - 11:20 am:

We are currently using Dionex DX500 for low level anion analysis (0.5 to 5.0 ppb Cl, NO3, & SO4). We have had a lot of trouble lately with the AG4A and AS4A 2mm columns from Dionex. We use a AG4A as a concentrator column and it may last us 4 to 5 days before we have to replace it. The machines seem to have to be recalabrated daily because of reproducibility. Has anyone else been having problems with any Dionex products? Any information would be helpful.

------------------------------------------------------------------------------------------------------------

By Chris Pohl on Thursday, February 27, 2003 - 01:18 pm:

Charles,

What specifically is the failure mode? Degradation of peak shape, loss in retention, incomplete recovery of early eluting anions, etc.? I need a bit more before I can help you but you can be certain we aren't hearing similar complaints regarding the application you describe. Concentration problems are typically connected to matrix issues. Often, I'm told the "matrix" has not changed. Nonetheless, usually the root cause is a minor change in the sample pH. Does your sample contain amine additives? Is there any change in the level or type of amine additive?

Regarding reproducibility, your reported issue with reproducibility is very uncommon. What is the magnitude of your calibration change from day to day? What method are you using for loading the concentrator? Is the reproducibility problem also observed with fixed loop injection?

------------------------------------------------------------------------------------------------------------

By Charles Wilson on Thursday, February 27, 2003 - 02:17 pm:

Chris,

It is a loss of peak height. We load the concentrator using an Alltech 426 HPLC pump. The sample that we mainly run on the column a very pure water sample (conductivity of sample is ~ 0.09 uS). This analysis is used to measure the anions in Reactor coolant water from a boiling water reactor. We check a 2ppb standard daily and the chloride and sulfate peaks seems to be most affected dropping usually 0.2 ppb a day. As far as a fixed loop system we were using AS12 columns (I think that what they were) but then the sulfate peak on those just seem to disappear in a day or two after new sets of columns were put in. This was using NaOH eluent and a Helium blank on the eluent. That is when we switched to the current method of concentration the sample but it is getting expensive with all of the concentrator column changeouts. The pH of the samples should not be changing hardly any at all at the conductivity that the samples are running at. IF you have any ideas or recomendations, I would like to hear them. If you need more information I will try to supply all that I can. Thanks,

Charles

------------------------------------------------------------------------------------------------------------

By Anonymous on Thursday, February 27, 2003 - 03:45 pm:

Dear Charles Wilson,
Regarding your recent message on the Ion Chromatography Forum,
Loss on peak height / area for SO4 could be due to lots of things. Sometimes the preconcentration column can get overloaded with other anions, e.g formate acetate etc. These can occupy sites that the SO4 would like to use. So if you continue to load preconcentration column that is already fully loaded with more formate/acetate, then your Cland SO4 peaks will be lowered. With some of the earlier Dionex analytical columns, the formate and acetate come out very early (too early) near the injection peak and so are difficult to evaluate/notice that you may have a formate/acetate problem.

Suggest you try the following.
1. Go back to a working system and use the fixed loop and test the analytical column first with higher standard to confirm IC is functioning OK. If not then thee is something fundamentelly wrong with your IC system.
If OK. next Try preconcentration again, but with standards only (not sample) to confirm IC is functioning OK
First start by reducing amount of standard (volume) you put onto the preconcentration column but use a higher concentrated standard.
If OK then go on to higher volume, but lower concentration of standard.
If this is OK try sample ... if not working, then there something in the sample interfering .... maybe problem is explained as above.

I am assuming that your standards are scrupulously clean and not contaminated.

------------------------------------------------------------------------------------------------------------

By Chris Pohl on Thursday, February 27, 2003 - 09:13 pm:

Charles

Assuming your sample doesn't contain boric acid as is the case in some reactor applications, I think we can rule out matrix problems. At the same time, the loss of peak height isn't necessarily any indication of a column problem even if replacing the column appears to "fix" the problem. The most likely culprit would seem to be the concentration protocol and/or the calibration method. It's safe to say that the guard columns can't be responsible for such a symptom, since they should last many months under normal conditions.

Concentration questions:
What flow rate do you use for concentration (flow rates over 1 ml/min on a 2mm concentartor can effect concentration efficiency)? Do you reverse concentrator flow after concentration? In what type of containers are the samples supplied? What is the peak height you observe under normal conditions? What concentration volume are you using? If you double and halve the concentration volume, do the peak heights and areas change accordingly (if not, you may be overloading the concentrator and the symptoms may be due to operation under non-equilibrium conditions)? Have you checked the volume of the concentrator effluent to see if it matches the expected volume? What is the backpressure on the pump during the concentrator step (a loading pressure under 500psi could cause pump reliability problems which can be corrected with added backpressure)? Are you alternating concentrators, using one for concentration while effluent from the other is being measured? If so, do the pressures of the two concentrators match? If you try a 100X eluent cleanup, does the concentrator show any improement?

Calibration questions:
What concentration is your standard (low ppb standards are generally unstable)? In what is it stored? How is it loaded? What is the loading volume? How often do you make up a fresh standard?

------------------------------------------------------------------------------------------------------------

By Charles Wilson on Friday, February 28, 2003 - 06:38 am:

Chris,
Our reactor coolant samples do not contain Boric acid, but they do have ~100 ppb Boron due to Tertairy Fission products. We run a flow of 1 ml/min during loading of concentrator for samples. They load a volume of 6 mls on one system and 8 mls on another. the loading pressure starts out at ~ 400 psi on a new column when it is first put in and after ~4 days the loading pressure is up to ~2000-2800 psi. Typically there are 10 samples run on the column in a day. Concentrator flow is not reversed. I have not tried to halve the concentration load or to double it but will try that. We do not have but the one concentrator on each system. We have tried to clean the column with 100x eluent but did not have any noticable results. We have two cal curves that we run on the system, one with a 0.5 and 2.0 ppb std, and one with a 1.0 and 5.0 ppb std. The standards are made daily from a 10 ppm vendor(VHG) standard. the flask that the samples/standards are run in, I beleive is a polycarbonate culture flask, 70 ml volume bent neck culture flask that was used in a RAM auto sampler.
AS for the other post, there is no evidence that we have seen of acetate or formate being in the Reactor coolant system.
Also at one time it seems that Dionex told us that running basically pure water samples through the column like we do due to the low impurities in the water that it actually causes the resin in the column to deteriorate(or swelling) in the concentrator causing the higher than normal pressure problems. Does that seems to make sense, it just strikes me as odd that pure water would do that.

------------------------------------------------------------------------------------------------------------

By Anonymous on Friday, February 28, 2003 - 09:48 am:

Just a suggestion, Another reasons why you could get low responses is that you have Fe in the eluent/sample. Iron (Fe) in the eluent/Sample could be depositing and sticking to the front end of suppressor. Try to regenerate the suppressor with EDTA or oxalic acid.
Fe on your suppressor will react with PO4 and/or SO4 and hence loose your response.
Have you got Fe in your sample / eluent ???
Beware, new brown glass bottles leach out Fe.

------------------------------------------------------------------------------------------------------------

By Chris Pohl on Friday, February 28, 2003 - 11:19 am:

Charles,

From the latest information, it looks like were getting closer to the cause of the problem. It is definitely not normal for the pressure to rise from 400 to 2000 psi after a few days of use. We need to identify the source of particulate causing the problem. One point to note is that generally, it is recommended that elution from a concentrator be performed in the reverse flow direction. However, I don't see why this would have anything to do with your pressure problem. Since you're seeing such a huge rise in pressure, have you ever tried reversing the flow direction to see what impact that has on the pressure? Also, is it possible there are suspended colloidal metal particles in your water? This seems unlikely but this would be one explanation for the pressure buildup. Otherwise, the source would have to be particulates from the pump or the sampler. Have you ever tried putting a filter in line to remove particulate? Upchurch makes several good ones for this purpose.

Regarding the suggestion from "Dionex" that running pure water to the resin causes the column to deteriorate (or swell), this is utter nonsense. I would appreciate it if you could e-mail me the name of the person that told you that (you can give my e-mail address by clicking on my name in this post).

------------------------------------------------------------------------------------------------------------

By Anonymous on Friday, March 19, 2004 - 04:34 pm:

Hi there!

I have a question, what if you cannot detect magnesium and calcium using IC system.When you do some calibration,you have to optimize and name the peaks with Mg and Ca and afterwards that you run your samples, those ions do not give any peaks at all but i am really certain that those cations are present in my sample.What could be the reasons behind? Another question is that, Does the concentration of eluent affect the detection of cations and anions? In what way?
Thanks! I hope to recieve your response as soon as possible. More power!

------------------------------------------------------------------------------------------------------------

By Chris Pohl on Wednesday, March 24, 2004 - 02:25 pm:

There are a number of possible explanations for your observed problem. Perhaps you can supply a bit more information. For example, do you observe this "lack of response" for standards injected immediately after the samples in question? What is at pH of your sample? What is at pH of your standard? What analytical system are you using (i.e. column, eluent and if relevant, suppressor)? One common cause of this problem is sample induced ion exchange retention in the sample introduction system. Most commonly, this is caused by contamination of the autosampler and/or sample loop by "life forms". You might want to try replacing your sample loop to see if this helps or try adjusting the sample pH to see if this makes any difference.