Analysis of sugars

[Rndirk]

Hello,

Our lab separates and quantifies sugars & polyols in a large variety of food matrices by GC-FID. This analysis is very often performed here and runs decent, but - correct me if i'm wrong - i have the feeling this method is really old and it can be done easier. I summarize our current method below. This is for monosaccharides, disaccharides and some polyols:

- Water extraction
- Deproteinization by carrez reagents
- Lyofillisation (calibration standards start here)
- Derivatization to oximes
- Derivatization to TMS
- LLE to i-octane
- Injection GC-FID

I'm looking for ideas, experiences how other labs do it. In the past HPLC-RI has been tried, but since we run any kind of food here, interferences were too common. Any experiences with LC-MS (HILIC?)?

[HPLC chemist]

Sugars analysis were initially performed as you described by GC (back in the 50's). But HPLC has become the method of choice. It depends on the sugars you are analysing and expected concentration. See ICUMSA (International Commission of Uniform Methods of Sugar Analysis) methods.

For starches and simple sugars it was RI detection and a Divinyl Benzene Styrene (DVBS) column in the Calcium form using water at 85 Celsius (Waters Sugarpak). The sample was prepared at 10% dry substance. This method was good for jams, jellies... after cleanup with a C18 Seppak.

Then it became Biorad Aminex at 210 nm (everything absorbs) with an organic and water mobile phase.

Most sugars can be analysed by an amino column (Aminex, which was the 1st HILIC column) with an MS detector but you get everything else. Thus, it depends on your sample matrix.

[Rndirk]

Thanks for the information.

I'm afraid universal detectors (RI, UV at short wavelengths,..) are not an option for us due to interferences we can not predict. I can see it's an easy option if the matrix doesn't change a lot. But here, we have to quantify the different sugars in anything really: beverages, milk, pie, fruits,...

I think LC-MS/MS is our only option? Or would post-column derivatization and fluorescence detection provide the selectivity we need?

In terms of GC, I would expect derivatization methods have evolved over the years. However, most methods i find are pretty similar to what i posted. Is that right or should I look further?

[MSCHemist]

I do sugars occasionally as well. I do PAAN derivatives though.

[MSCHemist]

I do sugars occasionally as well. I do PAAN derivatives though.

Dry in vac oven (dirty samples can be scavenged with hexanes first)
200ul hydroxylamine in pyridine
90 deg C for 1 hour
cool
200ul of acetic anhydride
90 deg C for one hour
cool
1ml water 0.5ml toluene
inject toluene

Fructose is a bit iffy as aldoses are converted to peracetylated aldonitrile and ketoses give a lower yield of cis and trans oxime peracetate. However, I get R squared with 2-3 9's in the cal and enough signal to quantitate. I can get up through the disacharides. If I got a higher temp column I could probably get trisacharides.

ITSD is aldonitol/ribitol

[HPLC chemist]

You can use FLD (fluorescence) after derivatization. Sounds like your sample matrix is 'very' challenging. MS/MS is very information rich and you may detect other things!

[HPLC chemist]

For extraction get a homogenizer and a 'Thompson' food paddler. Also be sure to spike samples and use enzymes a-amylase and isomerase to differentiate starch and maltose peaks.

[HPLC chemist]

Also, I would also specialize. You can't be all things to all men (I tried that in a previous life)! So you may want to do just simple sugars found (added) in juices, or jams, or jellies. So you may want to write a business plan and marketing plan.

[Rndirk]

MS/MS is very information rich and you may detect other things!

I'd use triple quad not to get a lot of information, but for the selectivity it provides. However, i don't have experience with how well sugars do in ESI/APCI, and i can't predict matrix effects. Maybe someone here does...

For extraction get a homogenizer and a 'Thompson' food paddler. Also be sure to spike samples and use enzymes a-amylase and isomerase to differentiate starch and maltose peaks.

Why would i want to cleave starch in my sample to maltose I'd try to get rid of starch before injection, since the method is for mono -and dissacharides it shouldn't be that hard.

Also, I would also specialize. You can't be all things to all men (I tried that in a previous life)! So you may want to do just simple sugars found (added) in juices, or jams, or jellies. So you may want to write a business plan and marketing plan.

I like this one but it's not my company

[HPLC chemist]

Based on my previous comments I presume you were determining the amount of starch, dextrins...since they are major components in foods. In addition, this method (HILIC/MS/MS) will be 'very' powerful. You may be the only lab in North America that can do this!

[Karen01]

However, i don't have experience with how well sugars do in ESI/APCI,

Negative mode with Chloride adduct works OK.