I am converting Fructose into HMF, Levulinic acid and formic acid are possible side products. I analyze the data via HPLC, bio rad column, I got nice results for some reaction set. But when I expect to have a very high HMF yield I face the problem of very high surface area under the peak of HMF which results in unlogical value for the yield which may exceed the value of conversion of fructose. Kindly note that I diluted the sample enough to be in the safe margin of my calibration curve.
I tried to figure out the problem, since fronting peak shape for those experiments with high yield could be the signal of our column loading capacity at its upper limit. They suggested to test the co-elution which we already did, and there was no problem with that.
is it possible that side products were shifted to HMF retention time? Shall I change the flow rate?
Your help is highly appreciated to solve this problem?
Note: Humins is a wide range of products, and it could be derived from fructose or from HMF so it could have several peaks, and for that, we need a wide range of standards ( if it is available).
FLOW Rate adjustment is NOT used to improve peak separation.
Peak shape fronting is usually an indication of column overload (or injection using a solution that is stronger than the mobile phase, both of which need to be corrected).
Good peak resolution and symmetrical peaks shapes are required for accurate HPLC integration (and therefore, accurate quantitation).
Co-elution must be considered and the method may need to be re-developed.
If the separation meets good chromatography fundamentals (IOW: K primes for each peak are good, peak shapes are excellent, stable baseline, etc), then reduce sample concentrations to work at lower concentrations relative to the column's capacity.
- "HPLC PEAK Fronting and Tailing, Common Reasons For It"; https://hplctips.blogspot.com/2019/10/hplc-peak-fronting-and-tailing-common.html
William Letter Many thanks for your detailed answer. It is highly appreciated.
I will try to reduce the concentration.
Thank you again
What is your injection volume? What solvent you are using to dilute the sample? What is your initial gradient? What wavelength you are using? Is it or close to the lambda max?
Always use mobile phase to make your last dilution before injection so it is similar to your mobile phase.
Navid J. Ayon
The injection volume for each sample was 20 μL;
Aminex HPX-87H ion-exclusion column (7.8 × 300 mm, Bio-Rad);
the mobile phase is water containing 0.005 M sulfuric acid; the flow rate was 0.6 mL/min, and the column and refractive index detector (RID) temperature was set at 30 °C
The injection solution and volume are fine for that column, but it is the sample concentration that is key here... and you did not provide this value (but we can skip this for now, please read on).
- Aminex HPX-87H ion-exclusion column (7.8 × 300 mm, Bio-Rad);
Well there is the problem. You are using ion exclusion (Aminex column) for this application. Those columns are easily overloaded. It is so important to filter samples before injection AND to run the most dilute solutions possible (enough that you can of course still detect them). However, why do it this way when there are much easier HPLC methods? Try a RP column (C18) with a simple acidified mobile phase being sure to keep everything nice and warm (Column temperature 30 to 45C. Try different temps to see what works best on your system).
BTW: I have consulted on a lot of "honey-bee" sugar analysis related projects in the last few years and see client's using outdated and difficult methods such as you describe. A quick keyword search on the web would reveal many better quality methods for sugars and organic acids, esp for 5-hydroxymethylfurfural (HMF). As I make my living as a professional scientific consultant, I can not say much more for free, but surely you will find better options with a few minutes of searching.
- Badges
- Science method