Polverone
Now celebrating 21 years of madness
Posts: 3186
Registered: 19-5-2002
Location: The Sunny Pacific Northwest
Member Is Offline
Mood: Waiting for spring
|
|
solvent system for TLC separation of coffee extract?
Well, it's that exciting time of year when students get to pick their own projects ("dig their own graves" in chem lab. The group I'm working with decided that it would be interesting to
compare the caffeine content of regular and decaffeinated coffees, both traditional and instant. And, just because we're oh-so-bloody clever and
thought we could save time over brewing the coffee and doing a liquid/liquid extraction, we did a liquid/solid extraction using the ground coffee,
dichloromethane, and a small amount of sodium carbonate solution, according to a protocol that we had on hand for solid caffeine extractions. We set
that up last week and came into lab this week, ready to filter our DCM extractions and quantify the caffeine from each of our 8 samples.
We were going to measure the caffeine content with ultraviolet spectrophotometry. Caffeine should be the only thing in our extracts that absorbs
around 234 nm. But, unlike our reference samples, the extracts we produced were disturbingly colorful. We had red, yellow, orange, even green DCM
solutions of who-knows-what mixed in with the caffeine we were after. Nobody's had this problem before - but nobody before treated coffee as a
solid. They all did solid extractions of chocolate or liquid extractions of beverages. Well maybe the colorful junk won't absorb in the UV.... oh
wait, it does. We have a big ol' plateau of absorption from about 220 nm up to 400.
Well maybe we can decolorize the extracts if we wash them with water... no, it seems not, but we sure can make a nice emulsion. Break the emulsion
with brine... and we still have lovely orange DCM (and don't forget the amusement of losing half of one sample as a student discovers firsthand
the pressure that builds up in a sep funnel when solvents are mixed).
What about decolorizing charcoal? Nope, turns out that's useless too.
Well what about separating the components with TLC to get an idea of where the caffeine lands, then separating larger samples with column
chromatography to quantify the caffeine?! (At this point, we had spent about 4 hours on what we expected to take about 45 minutes.) Professor guesses
we should try a half and half mixture of ethanol and DCM for the TLC plates. Spot, spot, spot... (repeat x100)... spot, spot, okay, drop the plates
into the solvent. Wait 15 more minutes, dry plates, whip out the UV light... and look at those lovely streaks! It's like a phosphorescent slug
climbed across our plates. Bands? We don't need no stinkin' bands! We'll just circle the whole plate and say "here be coffee
components!"
Yes, this is much more a steam-venting exercise as anything. But I am still wondering: what's a good solvent system for chromatographic
separation of caffeine, oils from coffee beans, and miscellaneous botanical pigments? I'd never done TLC before and wasn't expecting to,
until the chain of amusing disasters befell us. We're going back in on Tuesday to slay (or be devoured by) the dragon once more.
|
|
|
Lugh
Hazard to Self
Posts: 66
Registered: 29-3-2003
Member Is Offline
Mood: No Mood
|
|
you could try 30-40% MeOH in water. it works well for the separation of a lot of compounds using TLC and HPLC.
It might be easier to use HPLC run with a caffine solution as a standard, it takes a lot less time than TLC once you have it set up and dead niffty
for quantitative analysis
|
|
|
Nick F
Hazard to Others
Posts: 439
Registered: 7-9-2002
Member Is Offline
Mood: No Mood
|
|
Haha, sounds like you've been having a lot of fun!!
I've extracted caffeine from coffee a few times, but from strong liquid coffee. The DCM has always been coloured, but only lightly IIRC. Sort of
pale orangish colour. Might be worth noting that the DCM used by me was crude, extracted from paint stripper, and still contained some sort of plastic
stuff in solution, I found out later.
Very surprised that the charcoal didn't work, in my experience pet-shop charcoal is quite excellent at decolourising plant extractions
:confused:.
It's a pity caffeine doesn't come out well with an A/B extraction, that would really help clean it up. It does work, but since caffeine is a
pathetic base it is not as effective as on some alkaloids.
"Bands? We don't need no stinkin' bands! We'll just circle the whole plate and say "here be coffee components!"
"Well maybe we can decolorize the extracts if we wash them with water... no, it seems not, but we sure can make a nice emulsion"
- lol 
I hope your luck improves!
|
|
|
Polverone
Now celebrating 21 years of madness
Posts: 3186
Registered: 19-5-2002
Location: The Sunny Pacific Northwest
Member Is Offline
Mood: Waiting for spring
|
|
I hadn't worked with decolorizing charcoal before either. Maybe we didn't give the charcoal long enough to work. We went for 15 minutes of
magnetic stirring at room temperature, then filtered.
I would have found it all great fun if I had been prepared for it. But we were going off protocols used by earlier classes for very similar tasks, so
I figured it would be like most school labs: do exactly what you're told, get exactly the results you were expecting. Challenges? Difficulties?
Improvisation? Hey, those are all things for the home lab, not school! At school you're supposed to plug problem A into method B, collect your
data, do a few calculations, and collect your grade .
After doing a bit more reading about TLC I'm pretty excited about it. This summer I'd like to make my own plates and try running various
mixtures at home. I even realized that I had an Amateur Scientist article about TLC up on my website (http://bcis.pacificu.edu/~polverone/chromatography/1976-02-b...).
|
|
|
I am a fish
undersea enforcer
Posts: 600
Registered: 16-1-2003
Location: Bath, United Kingdom
Member Is Offline
Mood: Ichthyoidal
|
|
This is just a guess:
The undesirable uv-absorbers in your DCM solution are substances that aren't soluble in water. If you had brewed the coffee first, these
wouldn't have dissolved in the water used and so wouldn't have ended up in your final solution.
Try extracting the solute with water and see what happens.
1f `/0u (4|\\| |234d 7|-|15, `/0u |234||`/ |\\|33d 70 937 0u7 /\\/\\0|23.
|
|
|
Nick F
Hazard to Others
Posts: 439
Registered: 7-9-2002
Member Is Offline
Mood: No Mood
|
|
I think your best bet would be to shake up the DCM extracts with portions of ~1M HCl (emulsion forming is good this time, it'll help to protonate
and dissolve the caffeine), combine the aqueous washings after letting layers seperate (may be time consuming  ), basify with NaOH to pH 9-10 ish, shake up with portions of DCM, combine the DCM layers and evapourate. A/B
extraction style. The method that you thought you would save time by not doing .
|
|
|
|
![[*]](images/xpblue/default_icon.gif){kind=icon}
