Flavanoid sample clean up

[FragranceChem]

I'm working on an aqueous extraction of some various plants, and analyzing them for flavanoid content, and I'm having a similar issue to a previous post linked below:

viewtopic.php?f=1&t=12107&hilit=bread+extraction

I will also provide my elution conditions in detail with a chromatogram, but it's a pretty standard gradient elution consisting of water and methanol on a 150mm C18 5um waters sunfire column.

I will post my own picture tomorrow. I understand that water is not an ideal extraction solvent, and that I would be getting a lot of starches/polymers, but that is not going to change. I was wondering if someone could give me some similar advice or even let me know if SPE methods worked. Amylase was also mentioned as a possible solution; any extra info/experience with this would be welcome.

I understand that my overall flavanoid content will be low, and therefore any sample prep needs to be very thrifty in terms of not losing what little I might have. I have tried forcing the starches/polymers to crash out of solution using methanol, but then I run the risk of forcing out the glycoside versions of any flavanoids.

Thanks in advance.

[Gaetan Glauser]

Tannins are challenging to get rid of. I have never tried it myself, but phytochemists usually employ exclusion chromatography (e.g. Sephadex LH-20) or polyamide gel that can adsorb tannins. If I remember well, bovine serum albumin at 5% can also be used to precipitate them.

[Chris0000]

Hello

CPC (centrifugal partition chromatography) can be a succesfull method for (pre) cleaning crude extracts of natural compounds.

Best regards
Chris

[FragranceChem]

Sorry for the delay, this is a chromatogram of a gradient elution starting with 70% water, 30% methanol down to 48% A in 3.5 min, 39% A @ 7.5 minutes, to 33% A at 11.5 minutes, 60% A after 15.5 minutes and then held there for 10 minutes, and then back to 70% in 35 min. This chromatogram is at 350nm. The gradient is obviously set up to resolve the standards as nicely as possible, but I have tried changing the gradient and even tried isocratic separation with no better result.

http://i44.tinypic.com/2a636z9.png

Aside from the obvious quantitation issues the crap in the middle provides I'm also worried about seriously degrading the life of the column/pump. I've noticed my delta psi inching up as well as rythmic noise which can be seen at the end of the chromatogram.

I'm waiting to hear back from Waters about any known SPE application for tannin removal with SPE or SEC. Serum albumin sounds like it is worth taking the time, and is a much cheaper alternative than buying a new chromatography system. I've been given the green light on using propylene glycol as an extraction modifier so I will be trying that today.

[Hollow]

welcome to plant extracts...

if you're only interested in the flavonoid aglycas (and not the single glycosides), than you maybe can do a hydrolization with hydrochloric acid.

If you can't do the hydrolization you maybe could try something like (I haven't tried all of them myself, but search also for "polishing" process in food/wine industries):

- precipitation with Bentonit
- adsorption on skin powder
- precipitation with proteins (e.g. egg white = cf serum albumin)

SEC sounds also promising, whereas SPE needs carefull evaluation of the stationary phase. As this is also "only" chromatography, if the inferences don't separate on HPLC, they neither won't on a SPE of a similar stationary phase material.

[FragranceChem]

I should be getting some SPE samples tomorrow from Waters, and am cautiously optimistic. Propylene glycol seems to have lessened some of the tannin peaks. They are still prevalent, but it has allowed me to separate the injection much more fully with a different gradient.

I have a follow up question to my ideas regarding SEC, but it's not a strictly sample prep question. I'll leave it up to the admin whether it warrants being moved into a new post.

I was wondering what the theoretical implications of running columns in series are in terms of the pressure actually felt across the individual columns. I was wondering if running an SEC column in series with my current column could help polish up any the chromatogram. My line of thinking was that an SEC column either before or after the regular column would push any large polyphenols to elute quicker relative to the small flavonoids. The flavonoids themselves would not be affected except to maybe lose a little resolution because of longer retention times. An SEC column doesn't have the pressure capability of a traditional column.

Any engineering minded people out there who could explain how pressure is actually distributed or "felt" by the system?

[FragranceChem]

I made some new runs, and I'm having some moderate success, but I'm still getting some pretty miserable signal to noises ratios.

http://i42.tinypic.com/23rwd8y.png

http://i40.tinypic.com/1hz6o8.png

Same setup as before, the gradient has been modified, and A is now 0.1% acetic acid which I figured would improve peak shape. Any more ideas?

[Hollow]

maybe you just have to live with this baseline bump...

maybe the article below could also be of help.
I've successfully tried it for some other components.
For the reason of matching sample solvent and initial eluent, I've incorportated an evap and redissolutionstep. Maybe depends on your system.

Novel isolation of phytochemical compositions by phase transition extraction with acetonitrile
Journal of Separation Science, 2011, 34/3, pages 347–353
DOI: 10.1002/jssc.201000658

If you're only interessted in Rutin, than maybe you could try isocratic with acidified water and about 15-22%ACN, Temp 30-35°C.
This could decrease the "notice" of the underground as your comp will elute on the descending tail of this bump.

As for the column life, we usually add a column flush step (to >95%ACN) at the end of each injection, and do sample filtration with 0.45µm (or at least centrifuge 5min@20000g). (UPLC 0.2µm)

Doing so we rarely have any problems with column lifetime, normaly >600 Inj without problems