Injections of Aqueous Solutions into a GC

[mattmullaney]

Good Morning,

Suppose I find myself in the situation where I'm compelled to inject aqueous solutions into a GC-FID or GC-MS. I'm not allowed to change out the solvent that dissolves the sample, but at least I'm able to use a good bonded phase such that bleed should not be a problem. What measures must I consider to best ensure success using such experimental conditions? I should add also, I'd be wanting to do splitless injection, which I can only imagine would make things harder...though maybe not as hard as cold-on column injection.

My thanks to all for your consideration and advice, in advance.

[MSCHemist]

Put a column union and at least 2M of retention gap and replace or rinse (column rinse kit) frequently. Don't inject more than 0.5ul and ideally less than that, no stong acid or base. Ideally I'd see if a headpsace technique would work to exclude nonvolatiles.

[Peter Apps]

It depends on what is in the water that you are interested in analysing, and what is in the water that you are not interested in.

Peter

[mattmullaney]

Understood, and my thanks to you both for your patience. The water is to be used as an extractant, taking after BPOG (Biophorum Operations Group) draft methodology to to assess water-extractable content in (polymeric) materials intended for use in BioProcess manufacture.

Should have added this in at first...my description of the problem was incomplete.

[Peter Apps]

Understood, and my thanks to you both for your patience. The water is to be used as an extractant, taking after BPOG (Biophorum Operations Group) draft methodology to to assess water-extractable content in (polymeric) materials intended for use in BioProcess manufacture.

Should have added this in at first...my description of the problem was incomplete.

And it is still incomplete; "water-extractable content" covers a huge range of chemistry from ammonia to proteins via sugars and salts. The majority of it is completely incompatible with GC.

Peter

[mattmullaney]

Hmm, I see, okay then. The Biophorum Operations Group is an organization is a collaboration of the leading biopharmaceuticals companies:

http://www.biophorum.com/

They work to create a standard for the analysis of single-use systems in clinical operations...for example, medical devices. Thus, the polymeric materials I refer to are such that could be used in medical devices, tubings, valves, containers such as IV bags and others.

ASTM is also working on standard methods for the analysis of such materials.

So then, the extractable content from these single-use materials is varied, ranges from metals to organic species that are extractable in organic solvents (plasticizers, curing agents, antioxidants and others). Organic extracts are easier to deal with via GC, agreed. Water extracts are a different model (perhaps more appropriate) model for the end use of some devices that have medical applications.

Please, see what you think, and my thanks again for your patience.

[Peter Apps]

Things that extract into water but that cannot be extracted from the water by organic solvents, or purge and trapped or headspaced are poor candidates for GC anyway - so the restriction that you must inject water does not make much analytical sense.

There is no single analytical technique that will cover everything that might extract into water from medical devices.

Peter

[mattmullaney]

Peter, I'm Very Much Agreed...there's no sense involved. [Sometimes, recommendations from organizations don't make sense.]

That said, what measures could one take if, given no choice in the matter of injecting water into a GC without the opportunity to switch solvents out for the injection (and with little hope of actually extracting anything to measure on the FID) take to have the instrument operate throughout a sequence of measurements?

Limiting the volume of water injected would seem to be one good idea...is there anything else? I'm also compelled to use splitless injection, which seems even more...questionable.

Understood also, "I'm not making the rules, just trying to follow them" is an often-used argument...I don't like it much myself.

And, yes, other analytical techniques (e.g. ICP-MS, HPLC and others) will be brought to bear upon these aqueous extracts.

[Peter Apps]

Keep the inlet temperature as low as possible and the column temperature high enough to stop the water condensing. If you are allowed to use a programmable inlet (PTV) then do a cool injection and vent the water vapour if you can do that without losing analytes. Packed columns are tougher than capillaries so if you do not need extreme sensitivity and high resolution go for packed columns. For some analytes a hydrophobic packing in the inlet liner allows you to do a version of adsorb - thermal desorb while injecting liquid water - you need a PTV for this. There are some derivatizations that work in water - MSchemist is our resident expert.

The best advice I can give is just to inject water under standard conditions, destroy several columns and MS filaments while generating no usable results and then present the outcome to whoever it is is insisting that you adopt analytically senseless methods.

Peter

[mattmullaney]

Peter,

If you're ever near Philadelphia, PA in the U.S., please let me know, and I'll buy you a good meal and several beverages of your choice! I graciously accept your advice! Sounds like I'm in for some stormy weather!

And I'll suggest to the BPOG that changing out solvents after performing (useless) water extractions may be advisable. I'm only a little guy, but one never knows unless one opens one's mouth.

[Peter Apps]

Thanks for the offer, I'll keep it in mind.

The water extraction has some logic behind it in that the devices will be exposed to aqueous fluids when in use, but it would make more sense to extract with something like Ringers solution or artificial urine or tears depending on where the device is going to end up.

Peter

[mattmullaney]

Peter, I can only agree with your last post as well, and my thanks again!

[MSCHemist]

Acutally some things that extract into water can extract into the more polar organic solvents. Example I was working on acrylamide and if you saturate with salt (MgSO4 and NaCl) and do multiple extraction it will go into 2-butanone, ethyl acetate, or acetonitrile and then be blown down depending on how volatile.

[MSCHemist]

One qualitative technique I use to look at polars when I get flavors for duplication is to extract into water, then dry then use a general purpose silylating reagent such as BSTFA+TMCS or MSTFA+TMCS. IT allows me to see a wide range of compounds qualitatively and many are in commercial MS libraries though I don't favor silylation for quantitation. I pretty much extract, evaporate off the water then add dry pyridine (I store with NaOH pellets) and add the reagent and throw it in a microwave at 20% power for 2 minutes. Then I inject it 10:1 split into a db5 column and take it all the way to 325 degrees. I can see everything from nonvolatile acids, amino acids, sugars, nucleotides (works for inosinate guanylate looks like it decomposes). Check out the metabolomics protocols. Silyation and methyl chloroformate are big with them.

[mattmullaney]

Hi MSCHemist,

My thanks for your advice as well. We're still at a really early stage in our work. While we have a reasonable separation of organic-based extracted compounds, we don't yet really know what many of them are. We did get some good news yesterday that one peak observed in LC-MS is likely the same species as one that I observe in GC-FID (and that others have observed in GC-MS), took a long while to find proper ionization conditions in GC-MS (CI using qTOF) to have the info to come to this realization.

We've a ways to go...these BPOG/ASTM guidances are also still very new--ASTM's are not yet published.