It is not good thing to inject starch in an HPLC column. Try the following - dilute your sample 1:2 or 1:3 with you mobile phase and then filter through 0,45um. You should precipitate starch. This led to a sensivity loss, but you may do the protein/starch precipitation in only one step. good luck.

Starch and other large polysaccharides will precipitate in 70 to 80% ethanol, after that you can inject on your system after centrifugation or filtration. This should also work in 70-80% acetonitrile which probably matches your mobile phase. If those solutions don't work, you can try carrez precipitation. The carrez solutions are available to purchase from your usual chemical suppliers, or you can make your own.

You should treat the sample with solutions: ZnSO4 12% and NaOH 0.5M with ratio =1:1, v/v, for example: 5ml ZnSO4 12% and 5ml NaOH 0.5 M, then dilute with H2O to 200ml. Good luck

Thanks, I will try bringing them up in mobile phase, we were centrifuging them and filtering the supernatant, so they looked really nice, but hopefully this will knock the soluble starch out, too.

Hi: my advice would be to treat the samples beforehand with warm 90% DMSO. That is a classic starch-solubilizer (Carpita and Kanabus 1987). I fully agree with previous posts -- you don't want to put a lot of polymeric material on the column. Bonus with DMSO -- should remove some pigments, too. This could be an advantage for certain types of samples.

If you are interested in polysaccharides, you can degrade starch by treating it with alpha-amylase, and wash the degraded glucose away for several times. If the sugar you are interested in oligosaccharide, you might as well precipitate polysaccharides like starch by 80% ethanol. In a word, it really depends on which component you are looking for.

It depends on the column-detetector system. If you are using Pb or Ca based columns and a refractive index detector and operates at high temperature there are alwys problems with protein/starch (?) gloging the colums (potato, fruits-etc). I suggest to extract the low molecular weight sugars with some water alcohol solution MeOH or EtOH 50% (with an appropriate internal standard). Then Ri-detctor can be quite OK. We use DIONEX-PAD (pulsed aperometric detection) PA1 column as a standard for potato and cabbage analysis of sugars and have more or less stopped using Refractive index (both aquous or organic solvent with different columns). I do not recomend to degrade the starch by amylase at all- then you will produce glucose and malto oligos that you can not separate from the sugars you are looking for in you sample.

I agree with all those comments, that you need to improve your sample preparation. The high pressure occurs when you run samples right? It didn't happen when you inject standard solutions?

Yeah, the standards are always fine, and the samples usually are, too, we precipitate the soluble proteins, gravity filter, centrifuge, and then syringe filter. We just ran into a sample with a very high level of soluble starch. Yesterday we ran crushed up pretzels after extracting in ethanol and water and it ran great. My concern was the solubility of the sugars, but we spiked the sample, too, and all the recoveries were between 82-102%!

What is your analyte(s) of interest? I would propose that you t need to optimise the sample prep and compare the protocols suggested by others. Food is a complex matrix so you will still expect some interferences during the run, in comparison to pure standard.

I would have expected better than 82% recovery. Are there specific sugars with low recovery? Lactose (if its one of your analytes) can be tricky as it is probably the least soluble amongst the typical "sugars". If you have poor recovery of lactose you can heat during extraction (60-80°C) which should help.

The easiest would be using capillary electrophoresis at high pH with direct UV detection and then no sample treatment is required, no even filtration.

Yeah, we generally trend low on maltose and lactose recoveries, everything else is usually right around 100.

OK then I would try an "extraction" by heating for 15-20 min. Once they are solubilized they will stay in solution so you can cool back to room temperature before continuing with your analysis, hopefully your recoveries will improve. Just be careful, if samples are acidic you will hydrolyse sucrose, in such cases you'll need to neutralise before heating.

Just easily wash the sample with 70-75% alcohol (MeOH or EtOH). After few minutes sugar-free sample will be ready to be injeted.

I agree with Mahdi Kadivar. My advice is to do the extraction procedure with 70 % EtOH, the sugars are soluble, but the starch is not. After that pass the sample through Sep-Pack cartrigde or just through 0.45 um and then inject to the HPLC.

As already suggested, I also think that ethanol (maybe 80 % v/v) should be an easy way to eliminating starch from the sample. To accelerate the polysaccharide precipitation, try in low temperature followed by centrifugation (10,000 g, for example). I think by doing this your sample containing sugars will be easier to analyze.

If you suspect the running buffer then add this to the sample and see what happens ? If a precipitate forms centrifuge it out and run the soluble portion - you may want to wash the pellet to improve recovery ?

I agree with Mandi Kadivar and Nadezhda. Precipitating starch with ethanl may be the best choice.

Also depending on the sugars you are analyzing for maybe treat with amylase to digest the starch.

I agree with Mandi Kadivar and Nadezhda. to eliminate the starch with the alcohol solution 75-80%. and also pass the sugar solution through 0.45micro membrane filter.

Separate the high molecular weight soluble starch using gel permeation chromotography. then use the sample for analysis after collecting the sample (after 3 bed volumes- need to standardize).

80% ethanol is a traditional partion for simple sugars numerous referances to this can be found as published by southgate etc, there may be some value also in using a solid phase cartridge to separate soluble proteins carrez solutions also can be used to precipitate these if needed. If the solution is highly coloured use of polyvinylpolypyrrolidone (pvpp) can be used for phenolic compounds and activated carbon for other colours.

In sugar analysis in order to get accurate resuts with better resulation and sensitivity, we have to pay more attention to our sample preparation process. There are many factors direclty affect to column pressure, and resulation. Protein, Fat, Salt and Polysaccharides. As a first step , sample should be defatted using Hexane. Then sample can be extracted using 50 % Ethanol/water solution (reduces protein and polysaccharides interferences ). As final clean up pass sample extract through C 18 SPE colmn. Filter extract with HPLC micro filter.

Thanks everyone, we have been extracting with about 75% ethanol on a steam bath, filtering and centrifuging, and this has worked great. We like to clean the samples up as much as we can, but we are also a commercial lab, so we are under pressure to keep cost down, so things like SPE cartridges are a luxury for us, haha.

80% ethanol is the typical cutoff used for monosaccharide separation from polysac recomended by southgate etc

you might want to try a carrez reagents cleanup in case for a protein issue if SPE is not an option.

if the solution is coloured try poly vynal poly pyrilidone PVPP (Check spelling) to precipitate tannins

I have a questions about all the replies:

Will adding ethanol effect the glucose measurement? For example, will glucose reacts with ethanol? Thanks!.

At room temperature, glucose will not react with ethanol. If the sugars to be analyzed include sucrose, the sucrose will be precipiated by ethanol.

Will ethanol also precipitate the dextrin out?? I have some hydrolysate samples which is at the beginning of starch hydrolysis, and there are sticky and viscous. Even after filtration with 0.22 um filter, the liquid is not as clear as water (kind of like jelly). I believe this is because a large amount of dextrin is there. I ran those samples on HPLC and the pressure was going up. We are afraid that this will damage the machine. What do you suggest to do with my samples?

ethanol will precipitate a lot of higher mw polysaccharides out of solution What precipitates and what does not depends on the fianl concentration of ethanol. Normally you would probably go to about 80% ethanol if you want to leave mono and disaccharides in solution. Of course longer oligosaccharides will also remain in solution. If your sample is very viscous, maybe you want to dilute it a bit before injecting on the system. High back pressure is unlikely to damage the instrument, but may be a problem for your column. Make sure you set the upper pressure limit to the limit of your column. What is the column and mobile phase you are using ??

Sean Christopher Austin Thank you for the reply! Appreciate that.

The mobile phase we used is sulfric acid (0.005M) and the column is HPX-87H.

My experiment is to do a enzymatic kinetic study of my crude starch sample. So I added enzyme to hydrolyze. The sugar I want to measure is glucose.

My question is: 1. how much dilution you would prefer to do? I think the higher dilution of ethanol, the higher error you will get.

2. Why 80% ethanol can leave mono and disaccharide in solution? Does 100% ethanol will effect glucose? I am trying to understand the mechanism behind this.

Hi Yeye

It's just a case of solubility. Higher mw polysaccharides are not soluble in ethanol, so the more ethanol that is present the more the higher mw polymers will come out of solution. Most of the simple sugars remain soluble in ethanol.

Have you boiled your starch to put it all in solution before starting your enzyme hydrolysis ?? Sounds to me like what you are experiencing is the starch gelatinizing and going in to solution..... like when you cook. If you put flour and water in a pan it is not viscous. As you heat the starch goes in to solution and everything gets more viscous. Of course in cooking you don't add any enzymes so it stays thick, in your case, once your enzyme starts to work the viscous solution will start to thin out again as the starch is broken down.

Use the ethanol precipitation - your starch will drop out of solution and you'll be left with something you can inject on the HPLC, and it probably won't need much dilution, but you might want to try a 50/50 dilution or something similar, just to make the ethanol peak as small as possible.

Sean Christopher Austin Hello Sean,

Thank for the reply! I appreciate your patience!!

Here is my experimental details: I added varied dosage of enzyme to crude starch solution, and plot the enzymatic kinetics. Temperature is 55 ℃ (optimal for enzyme), and time period is 48h. Every triplicate test tubes represent each time point, so I sacchaifized tubes.

I sampled at different time points in order to plot the kinetics at the very beginning (5, 10, 15, 20, 25, 30, 45, 60min). In order to deactivate the enzyme, I put the tubes into boiling water for 10min. I agree with you that it is gelatinization of unhydrolyzed starch part leading to the high viscosity. For the treatment with lower enzyme dosage, I observed that the viscosity of solution was higher, because more starch was left in solution. However, when starch is hydrolyzed completely (or close to completion) and even though I boiled it, there is no viscosity issue.

I am kind of not confident with my methodology now, since boiling will gelatinize starch, increasing viscosity. I know there are other ways to deactivate enzyme, like adding NaoH... But I am also concerned that adding other chemicals will add error, especially when you add large amount.....

Since I already finished hydrolysis, I would take your suggestions to modify the sample preparation method for HPLC measurement.

Another question following up here: Since I have two different samples, viscous (and turbid) and nonviscous. Do I need to do the same sample preparation for all of the samples (dilute with ethanol)? Should I keep the consistency for sample preparation even though those nonviscous samples have no problem with supernatant.

Thank you

Yeye

I would always recommend to keep everything the same - so yes - add ethanol to all samples (even if it might not be nescessary) then you always compare "like with like".

Boiling to deactivate the enzyme is fairly standard practise - I wouldn't change that.

It might have been interesting to test your enzyme on native starch (as you have done) but also on pre-gelatinized starch (i.e. boil up the starch so it's completely in solution, cool to 55 then add your enzyme)..... anyway, good luck with your experiments. It should work Ok after using ethanol to precipitate away the starch.

Sean