Carbohydrates with a DAD

[IsotopeC14]

I'm running an HP 1050 refurbished system with a vacuum degasser and Agilent DAD (G1306A). I am developing methods in order to rapidly analyze wine for small organic acids and sugars.

I very clearly see in the literature that nearly everyone uses RI for sugar detection, but I resolve them fairly well at 192-196 with my DAD. I was curious if there are detection issues which I'm not taking into account regarding variability with small peak intensity (Glucose for example has an area of 10 on my run, at 0.2g/L 20ul injection, Fructose has a *much* better response though). I'm using a classic acetonitrile mobile phase and get a complete run time in of about seven minutes. I'm using a restek column and restek pre-made sugar standard. The sucrose, maltose and lactose come out after the glucose and fructose and are of little interest as they shouldn't be present in wine above trace.

Many classic standard mobile phases consist of a phosphate buffer of varying intensity. We've tried this and the machine cried. It was squeaking at the pump seals and started getting very noisy. It took a full day of running water through the system and replacing some seals. Are there any other buffers that are "less" hard on the system that people are using to replace say a 50mM phosphate bufffer? It seems to salt out obscenely within our system. It may be a crappy refurb, but the sample injector odometer was around 300 when we got it. (He might have "rolled" it back, wink wink). We've switched to a dilute H2SO4 for most of our small organic acid runs and it seems to work fine, we are just thinking about expanding into other molecules of LC interest that are published in the literature and almost all of them are phosphate buffer based.

Ok, thanks for reading and the future responses!

[Mark Tracy]

Did you get the seal wash option? If you are running buffers, it is a good idea. It is an easy retrofit. You didn't mention the pH of the buffer, but if it is around 2, you can probably substitute dilute H3PO4 which is completely miscible with MeCN.

[Noser222]

Just to let you know, I've also seen Fructose have a higher absorbance at 190 than Glucose. I guess it is due to the time each spends in the cyclic vs. linear form.

[IsotopeC14]

I think we would be most inclined to go the phosphoric acid route with any new development. We have milli-Q water and feel pretty good about our baseline overall. The seal wash option we will keep in mind if we *have* to go the phosphate buffer route.

Thanks for the perspective on the Fructose, I'd thought that might be why there was a difference, my background is in more of a microbiology/genetics background and I figured the sugars would probably work that way. LC is new to me, but I catch on preetty quick.

On a side note, we are messing around with a wine sample today here in our lab. We are running a restek "Ultra Amino" column. It is a 3uM particle size with a 100A pore size, mobile phase is 75% ACN. We have tried two different column extractions, one sample with a C18 separation column (restek) then extracted on a SAX column (restek), and one without the C18 column. We did not extract the acid phase from the SAX column since we only wanted to look at the sugar (water) fraction.

There is an awful lot of compounds that appear visible at 192/196. Some of them are also visible at 210, but not all. The glucose and fructose line up where they should be compared to the standards, but we see many other compounds that we'd not expected at 192. I'm figuring that we will probably need to send off a sample at get some LC-MS done on them. Anyone know of someone who uses either this type of column or similar and could also do the MS side of things for us? We are in the Pacific Northwest and would be happy to pay for the time/labor/reagents and such. Harvest is coming soon and we'd very much like to have an idea of what our capabilities will be for this coming year on the "sugar" side of things. I'm not sure exactly what we are seeing, but I'm guessing that they might even be amino acids since I don't think that they will bind to the SAX column terribly well. We may need to get a different prep column in the grand scheme of things, I'm open to all kinds of suggestions here!

[Uwe Neue]

Anything very polar will elute with the sugars from the amino column. There could be some salts even...

I assume that you did the sample cleanup on the C18 column in "pass through" mode. With other words, you activate the C18 as usual (MeOH, H2O), then pass your wine through. The stuff that breakes through then goes onto the (prewashed and activated) ion-exchanger. You can do still better by combining an anion-exchanger and a cation exchanger, since sugars are not retained by either.

Such steps can remove everything except material that is polar and neutral, like your sugars. Actually, this can be all done in one step with the right configuration of your SPE devices. With the right device, you can connect them in series and do everything at once. I personally would use Oasis HLB, Oasis MAX and Oasis MCX for this procedure.

[IsotopeC14]

I will order some more columns today and see how that works. I looked over the Oasis SPE site and found it to be quite interesting. I will likely just order similar columns from my current vendor as we already have an account and such. Since I'm in the private sector shielded from purchase-order people I don't get the chance to order things immediately from a new vendor, has to wait a week or so for "processing".

[Uwe Neue]

Look at the Oasis methods book. There is a method for acrylamide. My suggestion for the cleanup of your sugars was modeled after the work of my colleague for acrylamide.

[Mark Tracy]

Wine is complex. Sugars, organic acids, amino acids, polyphenolics, sulfites, and the list goes on. We have a sample of white wine by reversed-phase at 210 nm: http://www1.dionex.com/en-us/acclaim_li ... 28431.html We use the OnGuard II-P cleanup cartridge that is specifically designed to remove pigments from fruit juice and wine before anion exchange analysis because the red pigments will severely foul the column.

[IsotopeC14]

Here's an acid C-gram that we made from a Pinot Noir using a gradient program and some mild tweaking. We know that the large peak right before the acetic is lactic, but it's not in the calibration curve on this run. I followed the "sticky" on this and have never posted an IMG to a forum before so hopefully this works and isn't too big!

[Bryan Evans]

Increasing the column length and column temperature
will give you sharper peaks for sugars (as was done on Unison UK-Amino):

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

RP for organic acids is typically done under low pH environment / highly aqueous mobile phase conditions. Here are organic acids analyzed on
Unison UK-C18 and Cadenza CL-C18:

http://www.imtakt.com/TecInfo/TI348E.pdf

[IsotopeC14]

One of our biggest issues is really throughput. The current state of wine chemical analysis is that the majority of compounds are analyzed spectrophotometrically with an enzyme reaction kit. There are many sources of error regarding enzymatic analysis of wine. It is the industry standard and is limited since you can only get one compound at a time. Specs are somewhat variable in detection limit, and we've had more than a couple of natural random user errors that resulted in exploring more sensitive and reliable technology. Ultimately I need the analysis power of HPLC-MS/GC-MS on a refurbished HPLC budget with an analysis time of less than 8 minutes per sample. I think the very sharp C-grams are very attractive, but in the field we have to weigh great chromotography to analysis speed. Although a 30 minute sugar run looks great, we need to be able to get all of them done within a day (40 or so runs) in under 8 hours. The only thing that I'm paranoid about after seeing some C-grams is that galactose is awfully close to glucose in a 30 minute run, I don't recall reading that it would be in wine though. I really didn't like seeing the formic acid so close to the Tartaric, but theoretically it shouldn't be in wine either. It might be in jug wine but not the good stuff! I do appreciate all of the response, and I will certainly look into the columns that have been suggested!

[Bryan Evans]

Hi Isotope C14 -

Thank you for your response. A couple of things:
1. Pino Noir Chromatogram - I think that sample cleanup still needs to be evaluated. The big hump in the middle most likely means something is co-eluting with the peaks of interest. Also, you might be able to sharpen up your peaks by lowering the pH of mobile phase and / or sample solution.

2. Organic Acids chromatogram - It's very difficult to separate tartaric acid and formic acid on an ODS phase. TA and FA are not very well retained, and you need extra plates to achieve baseline resolution.

The good news is, if you are not worried about TA and FA, and if you're compounds of interest have a higher alpha, you can decrease your column length (and hence run time).

3. Sachharides chromatogram - Again, glucose and galactose are the critical pair. Better separation can be achieved by increasing k' (increasing organic since this is normal phase). But if galactose/glucose are not of interest - you can again decrease column length (and run time).

Side note - full baseline separation of glucose / galactose are below:
http://www.imtakt.com/TecInfo/TI322E.pdf

[IsotopeC14]

Thanks so much for your response Bryan, I hadn't thought about reducing the column length, we are already at a 150X4.6, and will look into some shorter columns (if they exist). I realized that I integrated that "hump" and shouldn't have. It's an artifact from the decreasing flow rate that we are running on this procedure, it is identical on our custom made standard that is completely flat if run isocratically. The *goal* was to get malic and acetic to have a larger area for the purposes of calibration. Our curve fits pretty nicely, all the way down to 0.025g/L which is dandy for a winery. There are critical acids in winemaking, and non critical ones. The critical ones are Malic and Acetic. In premium winemaking you want zero (or near zero) Malic Acid, and an Acetic acid concentration of less than 0.6g/L (variable). It is illegal to have I think more than 1.2g/L Acetic acid, so that is very critical. The Malic acid is converted to Lactic acid in the process, so the presence of Lactic is nice since it kind of acts as an internal control for the malic acid results in a C-gram. Right now we are running the acid side of this with a very dilute H2SO4 mobile phase (0.002N). As I state above in this thread, the phosphate buffer at pH 2.5 makes the HPLC sound like it is screaming in pain. The 0.002N phase pH is about 2.7. A secondary concern is the fumaric acid, which at high enough concentrations can thwart the malic acid to lactic acid conversion as stated in the literature. I noticed something interesting that when we ran the acetic acid kit (enzymatic) that we got a value 0.1g/L lower than what the HPLC detected. This was on a white wine, and either something co-eluted, or the kit is non-specifically reacting with something else resulting in an artificially low value for the spec method. I suppose in the grand scheme of things if HPLC companies could show that unfiltered wines (Wild and full of bacteria/yeast) will react variably with kits and not so with HPLC's that a new market may develop heh. The only issue is that wineries would start having to pay analytical chemists instead of "lab techs" lol (Don't forget a nice tip for me or a nice "freebie" column)

[Mark Tracy]

If you are going to be doing serious analysis of complex samples for organic acids, you should be using ion exchange chromatography. Reversed-phase is not going to do the job for you. If you can afford an ion chromatograph, the Dionex AS-11HC column will deliver the goods for you; see http://www1.dionex.com/en-us/webdocs/43 ... HC_V20.pdf page 32 for a complete method. A second choice might be the Acclaim Mixed-Mode WAX-1 http://www1.dionex.com/en-us/webdocs/48 ... olumns.pdf(we have not published an example of wine, but we can show you unpublished data if you want to see it).

Unfortunately your conflicting requirements for complex sample matrix, good separation, rapid analysis and low cost are all but impossible to satisfy. I know that in most wineries, the laboratory budget is spare change unless you work for E&J Gallo.

Winemaker: How do you make a small fortune in the wine business?
Chemist: Start with a large fortune.