Spike a Standard

[Jade.Barker]

Hi, Sorry for the long post, I wanted to be thorough.

We are a Non-regulated Lab. I have the authorization to change methods. Our HPLC:
Waters components: In-line Degasser, 1515 Pump, 717+ Auto sampler, 2414 RI Detector
Column IC-Pak Ion Exchange 7.8 x 150 with a Shodex SH-G Guard Column
Mobile phase: 0.5 mM H2SO4

We are currently running a 10 minute method; I have access to an equivalent 20 minute method as a last resort.

Our components of interest include (in order of elution):
Dextrin, Maltotrios, Maltose, and Glucose
Lactic Acid, Glycerol, and Acetic Acid
Ethanol

The Standard & Check Standard are acceptable.

A few samples have bad Chromatograms:
- Peaks shifting together early in retention (The maltotrios moves into Dex or Maltose)
- Peak Tailing or broadening early in retention (Maltotrios can be completely lost)
- Mystery peaks late in retention (In the long gap between Acetic Acid and Ethanol)

Here is my logic:
Since the Standards look good the issues must be sample related
It’s possible some samples are a stronger solvent than the mobile
Either known peaks or unknown peaks could be the problem

Decreasing Injection Volume?
Some samples have very low concentrations of the analyte of interest. (The standard has approximately 0.3%) Resolution priority should go to the acids, even if resolution suffers in the sugars. I could decrease the injection volume from 10 uL to 5 uL, but, I don’t want to lose the acid peaks into the baseline. Plus I still wouldn’t know which peak was the problem…

Spike the Standard?
So I would also like to "spike" different vials of my standard with some individual components to see if high concentrations of one component are interfering with other peaks. This would also identify if the shifting maltotrios really is maltotrios. It may be possible I have a totally different sugar in the area; I tried a sample from a different feed stock, sucrose co-eluted in my other sugars.

Questions:
1. Does my logic make sense?
2. What concentration of components would you start with when spiking a standard?
3. Is it acceptable to inject 10 uL first, and only re-inject problem samples at 5 uL?
4. Can you refer me to a resource for carbohydrate separation?

[JGK]

Have you the capacity to dilute one of the "problem" samples and still quantify it?

If so does diluting the sample improve the chromatography?

If the samples can still be quantified after a dilution it may not be necessary to alter the method.


Trying to improve separations could be problematical when using a system with an RI detector.

[Jade.Barker]

JGK,
Thank you for your quick reply.

I could dilute the problem sample for diagnostic purposes. However for normal laboratory processing it would not be practical given the number of samples we process and the technician availability.

I will take your advice and try it for diagnostic purposes on the problematic samples. I may be able to work that into tomorrow's schedule. I'll post back once I have tried.

Do you have any experience with adding a measured amount of a component to the standard to see the increase ratio causes interference with other peaks?

Also, I am very new to HPLC: Could you please explain more about how the RI detector is problematic for separation?

[JGK]

JGK,
Thank you for your quick reply.

I could dilute the problem sample for diagnostic purposes. However for normal laboratory processing it would not be practical given the number of samples we process and the technician availability.

I will take your advice and try it for diagnostic purposes on the problematic samples. I may be able to work that into tomorrow's schedule. I'll post back once I have tried.

Do you have any experience with adding a measured amount of a component to the standard to see the increase ratio causes interference with other peaks?

Also, I am very new to HPLC: Could you please explain more about how the RI detector is problematic for separation?

The RI detector is sensitive to:
Environmental factors (Temperature and Pressure)
Any minor change in Mobile Phase composition.

Gradient elution is not recommended when using RI.

[JGK]

JGK,
Thank you for your quick reply.

I could dilute the problem sample for diagnostic purposes. However for normal laboratory processing it would not be practical given the number of samples we process and the technician availability.

I will take your advice and try it for diagnostic purposes on the problematic samples. I may be able to work that into tomorrow's schedule. I'll post back once I have tried.

Do you have any experience with adding a measured amount of a component to the standard to see the increase ratio causes interference with other peaks? Not really as my approach would be the dilution of affected samples

Also, I am very new to HPLC: Could you please explain more about how the RI detector is problematic for separation?

The RI detector is sensitive to:
Environmental factors (Temperature and Pressure)
Any minor change in Mobile Phase composition.

Gradient elution is not recommended when using RI.

[rhaefe]

DiagnosticLab:

what is your separation temperature and flow rate? Are you applying the same conditions as in the Waters application note for the analysis of Ethanol in fermentation broths? In order to remove the lactic acid away from the glycerol you can lower the sulfuric acid concentration but I have to admit that the lowest concentration I tried was 0.5mmol/L. When using DI water as mobile phase the lactic acid elutes quite some time before the glycerol but the peak shape is terrible and quantification is impossible.
I had good success at 80°C separation temperature with flow rates between 0.6 to 1.2 minutes. The column I used was gel based (back pressure sensitive) in "standard" dimensions: 7.8x300mm.

How do you treat your samples prior top analysis?

[Jade.Barker]

Rhaefe, Thank you for your interest.

The column is 75°C and the flow is 1ml/min. We do follow the Waters Method “Fast HPLC Analysis for Fermentation Ethanol Processesâ€

[rhaefe]

Jade,

Ion-Exclusion columns can be run with 100% DI water. I was refering to such a run when I explained the problems encountered with lactic and acetic acids (broad peaks etc). The sulfuric acid protonates the organic acids so they can effectively be separated by ion-exclusion chromatography (a more detailed explanation about the separation mechanism can be found in textbooks). Preferable your sample should be made up or diluted with mobile phase.
Temperature control is important, protonation equilibria are temperature dependent and mobile phase viscosity (back pressure) are kept to a minimum at high temperatures. You might want to refer to your column data sheet to find out the maximum T, but I have used ion-exclusion columns up to 95°C.

Most companies offering ion-exclusion columns publish k tables or similar selectivity charts for different analytes. Examples can be found at Transgenomic's (www.transgenomic.com) or Hamilton's (www.hamiltoncompany.com) web sites.

I usually prepare mobile phase from standardized reagents (e.g.:
This is the easy way to prepare mobile phases...

[Jade.Barker]

Rhaefe,
Thank you for your continuing input.

I am *seriously* suspecting the lactic acid the problem when it gets above a certain level. I was planning on running some experiments today… But my column conditioning step is showing a peak at 5 mins I need to get my system nominal before I can try anything.

I do have the option to dilute ect for diagnostic purposes of this issue, but I’m afraid it is not going to be practical for regular operations. Two other people also operate this instrument. An added preparation step could create an un-acceptable level of variation. How do you manage varying levels of technician skill?

I am interested in reading up on back pressure issues. I noticed my pressure was approx 100 higher yesterday than it has been running at. Not a huge difference, but I want to keep my eye on it. Do you have a good reference for gel column/ backpressure issues?

You make a good point about checking the column documents for temp max, I will check my documents. Also the “k tablesâ€

[rhaefe]

I usually use 0.1 or 0.2 Normal sulfuric acid and dilute appropriately. Not sure if 0.5mmol/L sulfuric is available.
Gel columns are fragile so make sure to prevent excessive back pressure at all cost. Excessive pressures will cause the gel like stationary phase to fracture due to the mechanical stress) and the column is gone. Check your manual for maximum pressures. Also make sure that your column compartment and column has reached the set T before ramping up the flow rate.

I can recommend three books for ion-exchange chromatography:

Ion Chromatography, Paul Haddad and Peter Jackson, Journal of Chromatography Library, Vol 46 (1990). Chapter 7 (27 pages) deals exclusivley with ion-exclusion chromatography).

Ion-Chromatography, James Fritz, Doug Gjerde, Wiley-VCH

Handbook of Ion-Chromatography, Joachim & Tatjana Weiss, Wiley-VCH (2005)

I think the first one is out of print but you might be able to get it second hand. The Haddad and Weiss books are pretty expensive (around $400-500.00).

The book from Weiss is written from a more "Dionex-centric" point of view...

[Jade.Barker]

Robert,

Thank you for the book sugestions. I have forwarded the book list to my Manager - maybe it will be an eary X-mas for the lab?

Update: I spiked my Midland Ethanol Standard with Lactic Acid. The standard is normally 0.3% Lactic Acid - but real world samples can be much higher. Wow! At 1.5% the Glucose Peak gets a nasty shoulder. At 3.0% it is almost a split peak. It's really dramatic.

Now, I'm thinking of other components that might throw the ionic separation out of wack. I talked to our PhD, get this: Some of the fuel ethanol plants use Caustic Soda to clean almost everything... I bet that sodium is working it's own magic in other problem samples.... I can't wait to find out.

Thanks for the insights - keep em comming!

[rhaefe]

What's the pH of your samples? If they are basic you might want to add enough sulfuric acid to them to get the pH down to 2-4. otherwise the mobile phase may not be enough to completely protonate your organic acids in the sample.
Hope Christmas comes early this year!

[Jade.Barker]

Robert,

I took the pH of some real world samples: Many have a pH above 4.7... The mobile phase is usually right around 3.3. Is this enough of a difference to be the cause of a problem?