Shark-Fin peak from biomass samples

[jase1337]

As you can see in the HPLC chromatogram below, I am getting this giant peak eluting where xylose normally appears. This trace is from an acid hydrolyzed biomass sample generated using NREL LAPs.



I am using a Shodex SP-0810 column, with deashing and carbo-p micro guards. Ideally I'd like to visualize and quantitate as many sugars as possible, (xylobiose, cellobiose, glucose, xylose, galactose, arabinose, mannose). The only problem is this giant peak that tends to slide along the baseline interfering with any sort of repeatable quantification. I can get rid of the "shark-fin" if I constantly feed the HPLC new anionic guards in the deashing set, (pricey, as I'm only getting about 20-30 injections/guard). I've tried different sample clean-up methods, EtOH, ACN, MeOH, etc. ppt.'s, which do remove the shark-fin, but this yields in a huge loss of sugars. I've also tried batch treatment with various cation/anion resins, but these only decreased the width of the peak.

Sorry to be so long-winded, but I'm at a loss as to what to try next. Any help would be greatly appreciated.

Thanks!

[Noser222]

Are you using sulfuric acid for hydrolysis?

Are you using RI for detection?

When you say you tried "batch treatment" with various resins, do you mean solid-phase extraction?
If you have luck removing it with a deashing cartridge, that same phase should work for solid-phase extraction.

[philippem]

Hi,

to me it looks as the shark fin peak is overloaded.

have you tried to inject less sample (amount or concentration) .
Do you know which compound this shark fin peak stands for ?

Philippe

[Bryan Evans]

Hi Jase -

Just curious - why are you seeing a loss in recovery in sugars
when doing a sample cleanup with organic? Is it because the sugars
are not soluble in 100% organic (have you tried 90:10 organic:H20)?

Is your sample solution too dilute (can you evaporate over nitrogen?)

[Bryan Evans]

I asked about dissolving in organic because if you can -
you can separate your sugars using normal phase on
our Unison UK-Amino column (provided you have RI or ELSD:

http://www.imtakt.com/TecInfo/TI327E.pdf
http://www.imtakt.com/TecInfo/TI325E.pdf
http://www.imtakt.com/TecInfo/TI320E.pdf

[Bart]

Hello Jase,

Seems as though you have a difficult sample prep. Since the interfering peak obviously has some interaction with the column material, it will be difficult to separate it from the sugars using sample prep given the similar chemical properties. It is curious to me as to why you are able to remove it on-line with an anion guard, but not off-line with anionic materials. Do you know the difference between the on-line cartridge material versus the off-line anion materials? Also, when you mention the interfering peak "slides" along the baseline, do you mean it elutes at different times (indicating it is highly retained on the polymer, and then eluting later at various times depending on the time between injections)?

Have you tried to contact the NREL? We are actually working with them on this separation using our column technology to try to improve the separation. They are a good group to work with, but can be a little slow in response (government). I realize you are most likely constrained by their LAPs, but feel free to contact me directly since we might have some insight into solving the sample prep problem and we also have similar technology to the Shodex columns. We also have the NREL LAPs.

[jase1337]

To all:
Thanks for your replies!

First, let me answer some of the questions posted:

Noser222 - Yes, I am using H2SO4 for the acid hydrolysis, and an RI detector. By batch resin treatment, I am adding resin to my sample, a few mgs/mL, then pellet the resin and inject the supe.

philippem - If I load any less sample, I lose resolution on some of the sugars that are lower in quantity than others. Also, the obstructive peak is still overlapping some of the sugars I want to quantitate.

Bryan - I am not sure why I am losing sugars from the organic solvent treatment. After treatment with the organic, (I have tried diff. concentrations, and 1:1 with EtOH worked best as far as chromatographic resolution is concerned), I speed vac my samples, then bring them back up to original volume with diH2O.

Bart - I'm using Bio-Rad's deashing set (H+/CO3-). I have inquired to Bio-Rad about obtaining a resin that is most like that of the anioinic part of the deashing column, but the don't sell it in that form. The resin I did obtain from Bio-Rad, I converted to the CO3- from with carbonic acid, (per their SOP), which still did not work. The "shark-fin" will elute at different times if I continue to inject many samples (100-200+). We did contact NREL about helping us out, and they offered to send someone out...for $30,000! Needless to say we're trying to go at it on our own.

[Noser222]

I've only just begun using this type of column (Pb form), so I admit I don't have much experience with yoru problem, but is there a chance you have a lot of an organic acid in the sample? If so, perhaps you could try an amino phase sample prep.

[Bart]

Noser222 makes an excellent observation. The shark fin peak shape is exactly what you would expect to see for weak organic acids eluting on a carbohydrate analysis column. You can test this by running the same sample on an organic acids analysis column using dilute acid eluent (I'm assuming you have one for the other NREL LAP for the organic acids test). You can eliminate the sugars from interfering by using UV. If you get one of more sharp peaks, you've identified the contimination.

[jase1337]

I don't know if this helps, but I have injected blank acid hydrolyzed sample, (4% H2SO4, w/out biomass, neutralized with CaCO3 through a spin filter), and only the shark-fin popped up. This led us to believe that what were are seeing is soluble CaCO3.

I collected several fractions of the peak and passed it to our analytical dept. Based on NMR results, they said it looks like "long chained carboxylic acids of alcohols."

Based on this info, I am still unsure of how to remove these possible contaminants from my samples.

[Bruce Hamilton]

I'm not familiar with the particular column, but perhaps can suggest some ideas for you to try. Just to cheer you up, I think you have two problems - Inadequate sample preparation, and inconsistent chromatography. However, if you fix the first, the second might not be a significant problem.

You haven't given us details of the chromatography ( eg mobile phase, column temperature, sample solvent ( water? ), injection volume ).

The nmr data suggest that perhaps your neutralisation of the acid isn't complete or consistent, and perhaps sample doesn't completely dissolve. If the problem peaks from the samples are long chain organic acids, sample prep should be able to remove them, provided they are made soluble and available to the resin.

I'd be looking to perform a more vigorous neutralisation with pH monitoring to ensure all of the sample is well neutralised, and stays at the desired pH for at least an hour after. As you can lose sugars, I'd also be focussing carefully on any concentration step as well, sampling and testing subsamples during concentration to ensure pH and sugars are not changing. I'd also want to be certain the samples and standards are at the same pH, and staying in solution.

Do you have any problems with inconsistent standards, and what happens when you spike the samples with the standards - is the recovery good?. If you increase/decrease column temeprature, do the peaks change size?.

I suspect that you may accumulating some material on the guard or column, and your sample injection solvent is liberating a small amount, and carrying it through the system slowly. However, it's retention should still be consistent, so that suggests the mobile phase may not be able to buffer the compound correctly.

As your anionic guards can prevent the peak for a while, that suggests you should focus on the sample preparation step, particularly pH control and sample concentration/redissolution solvents. If that doesn't remove the peak, investigate the anionic cleanup, perhaps trying different resins, or even much larger capacity anion guards from another supplier.

Please keep having fun,

Bruce Hamilton

[Bart]

Is it possible you have some bicarbonate left in solution? Bicarbonate is retained on these types of columns. Have you tried an injection prior to neutralizing the blank? The guard column should protect the column from the acidic blank.

[jase1337]

Bruce -

You haven't given us details of the chromatography ( eg mobile phase, column temperature, sample solvent ( water? ), injection volume ).

Mobile Phase: HPLC grade water, 1mM NaN3
Column Temp: 80C
Injection Vol: 5uL
Run time: 35min
Sample Dilution (after neutralization): 5X

I'd also want to be certain the samples and standards are at the same pH, and staying in solution.

I have checked the initial pH, but not after an hour. I will look into this. Everything seems to stay in solution, I checked my plates after a few days. and did not notice any ppt.

Do you have any problems with inconsistent standards, and what happens when you spike the samples with the standards - is the recovery good?. If you increase/decrease column temeprature, do the peaks change size?.

I have run acid hydrolyzed standards and I am getting almost all of the sugars back after comparing what I'm putting in, to what I get back, provided the shark peak is not interferring with the standard. I haven't tried playing with the temp.


Bart - I have not tried to inject any samples that were not neutralized. I thought that this might harm my set-up as the pH is around 1.

[Bruce Hamilton]

I don't know the column, but probably would not use sodium azide in the mobile phase - unless Shodex recommend it. I never have used it, as it allegedly forms hydrazoic acid very quickly - just another potential complication in an analysis.

My limited understanding is that these columns will have fairly consistent retention times for sugars, but retention will vary much more for organic acids, even changing according to the acid concentration.

Given that you have no buffering ability in the mobile phase, and the peaks moves around, I suspect that your samples may be varying in pH, or carrying enough ions to cause a similar effect.

I assume that the mobile phase has to be pure water for optimal separation of sugars, in which case you should work to ensure that your samples are as clean and as consistent as possible.

My question about spiking a sample remains. what happens when add some standard to a sample, hydrolyse both, compare against results for hydrolysis of sample and standard alone?.

It sounds as though your guard columns are doing the job, so you have something to work from, and I would focus on sample preparation using a resin if you can show consistent pH from your current treatment.

Please keep having fun,

Bruce Hamilton

[Noser222]

Why sodium azide? Your column is in the lead form, so you should only be using a mobile phase of water or a lead salt.