Sucrose analysis

[miro2009]

Hello,
We have been asked to develop a quantitative HPLC method for sucrose analysis present in a finished biopharmaceutical product... and as all our work mainly depended on UV, we are going to purchase an RI detector module for our system as according to my limited experience for this analyte RI is the optimum way of detection. I have a lot of questions and screened a lot of literature about carb. analysis in general and feeling a little lost, and would appreciate if anyone has links to specific practical and to the point webpages on handling such technique and important tips of handlinf, and how it differs from normal UV HPLC work, for eg:

- I would like to know how sensetive it would be as our samples would have a conc of 30 mg/ml sucrose, will that be easily detectable and quantified?
- would any sample pre treatment be needed? it contains a glycosilated protein in a phosphate buffer, would this interfer with sucrose detection? if so, what is the general way to treat the sample "quantitatevely" without affecting sucrose accurate quantifiaction??
- I read a lot of column types could be used, but I"d prefer an isocratic RPC method for simplicity, after all the method is just for analysis of an excipent, so any column and company recommendations? and any method setup suggestions?

Awaiting your "practical" feedback to help me imagine this new field

[tom jupille]

If you use RI detection, the method will be isocratic (by definition!). You might want to consider one of the evaporative detectors (e.g., ELSD or CAD) to give a bit more flexibility by allowing gradients (albeit at the price of a more complex system)

30 mg/mL should not be a problem; in fact, you will probably have to dilute.

Sample cleanup depends on how much protein is there and how many samples you will be running. Some thoughts:

If the sample load isn't too heavy, you might get away with using a guard cartridge and "dilute 'n shoot" (you'll probably use up a lot of cartridges!). Precipitating the protein with something like ACN might work as well (and you'd end up with a solution that you could inject directly onto an amino bonded phase HILIC column.

[mardexis]

As Tom suggested... Elsd would give you more flexibility as you would be able to run gradient analysis which you can not do with RID. The RID probably will get you a little better linear range and you wouldn't have to supply nitrogen (consumable). A lot of people do sugars with ion-exclusion. The biorad 87H or similar Phenemenex Rezex column are slow but work well. With a 5 mmol sulfuric acid mobile phase and the 300 x 7.8 mm column and 0.6 mL/min the retention time for sucrose should be around 7.5 minutes. Sucrose is reported to degrade at 50C on column so run it at 30C. For sample prep I would typically dilute 1:1 with ACN to crash the protein then spin it down and inject supernatant. As I recall the range was something like low umolar to high mmolar linear range depending on injection volume. Good luck!

[SIELC_Tech]

I like ELSD much more than RI, linearity and sensitivity are much better if you consider all factors. You can do any type of gradient and inject large amount into the column. here is a method for sucrose and other sugars using Primesep S column and ELSD detection:
http://www.sielc.com/pdf/Sugars%20on%20 ... column.pdf

You can dilute your sample 10 times with ACN and inject it.

[HPLCCONSULT]

With such high concentrations of sucrose in your sample, RI detection may be fine and less expensive to run. If you have temperature under control with your RI detector, then you should be able to obtain a nice signal.

As noted earlier, ELSD works very well for sugars. Here is a link to a simple glucose concentration study using an HPLC method with ELSD detection for Glucose (not sucrose, though the result should be similar).

http://www.hplctools.com/HPLC_Glucose_R ... C_ELSD.pdf

[miro2009]

Thank you all for your replies, I'm afraid I still do have some questions, as I said this is the first time for us to work in this specific field (ie. detectors other than traditional UV detectors):

1- as I said we are going for RI detector as we already arranged the purchase for an RI Shimadzu module as we believe it is simpler than ELSD. You attached some links for ELSD based methods and columns, so I was wondering do you have further specific input for setting up an RI method (column type/name/brand preferably RPC, mobile phase composition preferably if ACN based....)?

2-the components of my samples is 18 ug/ml of a specific 30KDa protein, sucrose 30 mg/ml, and 0.1 sodium phosphate... so injecting the sample as it is wouldn't be recomended? what are the general rules in RI of interferences in detection?

3- I would also appreciate if you have any good links about how to effectively deal with RI detectors (for eg. how avoid baseline drifts noise, temperature stability...), and how different it is dealing with it compared to a UV detector.

Thank you for you coperation

[tom jupille]

setting up an RI method (column type/name/brand preferably RPC, mobile phase composition preferably if ACN based....)?

From the point of view of separation chemistry RI detectors and evaporative detectors are fairly similar. The big differences are:
- RI methods must be isocratic (evaporative detectors can be used with gradients)
- evaporative detectors are limited to use with volatile solvents/buffers/additives (RI detectors have essentially no limitation on mobile phase composition).

The reason you see more applications notes on evaporative detectors is that they are newer (RI detectors have been around since the late 60's), and so have a greater "show me that it works" factor.

the components of my samples is 18 ug/ml of a specific 30KDa protein, sucrose 30 mg/ml, and 0.1 sodium phosphate... so injecting the sample as it is wouldn't be recomended? what are the general rules in RI of interferences in detection?

You have almost 2000 times more sucrose than protein in there. I don't think the protein will matter. The high sodium level would be a problem for the "cation exchange" type packings (e.g., the Aminex HPX87 columns). That would tip the balance in favor of the amino bonded phase packings. Most of the "usual suspects" have dedicated carbohydrate columns of that type. Probably several of them will chime in here

I would also appreciate if you have any good links about how to effectively deal with RI detectors (for eg. how avoid baseline drifts noise, temperature stability...), and how different it is dealing with it compared to a UV detector.

Per my comment above, RI detectors have been around for so long that very little gets published on them these days. The biggest issue is temperature control; some years back I ran a "back of the envelope" estimate that the noise spec on an RI detector is equivalent to something like the RI change caused by a 10-4 degree C temperature change. Second biggest issue is flow and composition; my experience has been that it's better to use pre-mixed mobile phases (rather than mixing on-line). All of that said, at 30 mg/mL, you won't be pushing the sensitivity of the detector, so I wouldn't expect much in the way of problems.

[miro2009]

Dear Tom,
Thanks for the helpful reply, I would like to ask your opinion on column choice based on your reply below:

The high sodium level would be a problem for the "cation exchange" type packings (e.g., the Aminex HPX87 columns). That would tip the balance in favor of the amino bonded phase packings. Most of the "usual suspects" have dedicated carbohydrate columns of that type. Probably several of them will chime in here

I was originally thinking to buy one of Biorad Aminex columns, but after your input, and after screening reputable column suppliers, I came up with Tosoh TSKgel HILIC columns, they offer silica based resins bonded to amino (TSKgel NH2-100) and carbamoyl (Amide-80), the latter claimed to be more stable, more details in this link: http://www.separations.asia.tosohbiosci ... 6_0807.pdf

1- What do you think about these options?
2- would you suggest any other better alternatives (company/brand/column name)?
3- And would you say with these HILIC columns the sodium phosphate now makes no interference?

Please excuse my lack of "hands-on" experience in this specific field, and all your input is highly appreciated

[dorota]

There is also a paper on UV detection:
"Separation of fructose, glucose and sucrose in fruit by high performance liquid chromatography using UV detection at 190 nm" J. Food Agric. 1983, 34, 109-112

I use Nanosep filters (http://labfilters.pall.com/catalog/laboratory_20025.asp) to remove proteins from samples.

[tom jupille]

1- What do you think about these options?
2- would you suggest any other better alternatives (company/brand/column name)?
3- And would you say with these HILIC columns the sodium phosphate now makes no interference?

To take those in reverse order:
3. I don't expect the phosphate to be a problem with any of the HILIC methods.
2 & 1. Based on that link, the TOSOH column looks fine. Unless there's something else in the sample that you haven't told us about, pretty much any amino / HILIC column should work. If this were my problem, I would buy from a vendor that I have had good experiences with, subject to the condition that they could show an application note for sucrose on that column (the TOSOH column does) so that you don't have to "reinvent the wheel" in figuring out the conditions. I like to bunch my purchases with a small number reliable vendors because it gives me more leverage when/if problems occur.

[lmh]

Not wishing to put you off what sounds like a very reasonable approach, if you ever find yourself in need of a cheap and effective 2nd approach, there are good coupled enzyme assays for sucrose that can be carried out if you have access to a UV spec or plate-reader capable of working at 340nm (which I'd guess is fairly likely if you're also doing protein work).

[tom jupille]

You mean there's something other than chromatography??

I'm shocked! Shocked, I tell you, shocked!!!

[juddc]

Hi There -

A few small points that may help in making the decision between the RI and ELS(I hope):

1. The RI will take significantly longer to equilibrate than the ELS. I have, use, and love my old RI detector because it's stone reliable, but it can be frustrating if I am in a hurry. In a QC situation, some forethought may be needed to obtain optimum efficiency. What I often do if I'll be running my RI is set up my mobile phase the night before, run ~20-30 system volumes through to waste while purging the RI, then run the line from the purge outlet back to my mobile phase reservoir and let the system recirculate overnight at whatever flow rate I'll use for the analysis. In the morning, I just close the purge valve and we're pretty much ready to go. If you need speed, an ELS takes about as much time to equilibrate as a UV detector. Advantage: ELS

2. The ELS will have poorer peak area reproducibility from injection to injection than the RI. I've seen a peak area % RSD of 0.5% for 5 injections, but that's really good for an ELS. Typical is 1-2% and can be higher depending on your system. The RI will be about as good as your UV detector, assuming a stable baseline.
Advantage: RI

3. The ELS will require more maintenance and care than an RI and, as others have said, is more limited in the mobile phases that can be run through it. Lamps need replacement, nebulizers need cleaning, nitrogen generators need maintaining, etc. It's not a hassle in my opinion - my ELS has been very reliable and I'm happy with it, but my RI functions as well as it did when I bought it 15 years ago and has had NOTHING done to it in that time. Nada.
Advantage: RI.

4. An RI will be cheaper to buy and simpler to set up than an ELS. The ELS isn't difficult to set up, but you do need a souce of clean, dry compressed air for your nitrogen generator. For the RI, you need a bench and electricity.

With all of that said, having access to both, I'd probably opt for the ELS and a small HILIC column. Method development will be much faster (you can change conditons and not wait hours between injections) and your Qc lab won't be complaining about equilibration times.

Hope this helps!

Chris

[miro2009]

Thanks all for the replies, we will try to think about an ELS detector but obviously not in the near future! We just invested in an RI and flourescense detector, so we would have budget limitations... and to be honest my literature screening gives me the impression that RI was the better choice for our goal, as our work doesn't deal with complex sample mixtures of carbohydrates nor acids nor inorganics in general...

Back to RI, it seems I will have to make some trials and errors in the beginning, probably starting with the Tosoh column, but questions keep arising, mainly this very common observation:
all methods seem to recommend very elavated oven temperatures up to 80 degrees, to avoid split peaks. When I went through the specs of our new RI detector I found the following:
Temperature control of cell unit: 30°C to 60°C,
Operating temperature range: 4°C to 35°C

As we are still gaining experience with this new detection system, are the above detector specs normal? and can it be operated at a temperature different than that of the column oven?

[mardexis]

The Tosoh amide is my favorite column for this app. You might also try the new Waters
UHPLC amide phase if you want high throughput. I have heard that amino phases tend
to form a schiff base with the sugar and performance degrades pretty rapidly.

Not to worry about temperature control of the cell. I think. Check with your sales rep
but your RI must be able to both heat and cool to obtain the desired cell temperature.
The RI will control the temperature that you need in the cell. The column can be 80C
but the RI cell will be some other controlled temp. i.e. no need for both column and RI
to be set at the same temp.