GC, Water and etc.

[Roxanne42]

This particular group wishes to study an enzymatic-style reaction concerning the conversion of 2-butanol to 1-chlorobutan-2-ol. Their reactions would be conducted in water as a solvent, with the presence of metallic catalysts and NaCl (among others).

They were wondering if there was any convenient way to analyze their reactions with the GC... Water samples are obviously not recommended on most columns - are there any columns out there that can withstand water samples?

Evaporating the water to replace it by a solvent is obviously not feasible with the low-boiling point of some of the compounds of interest. We also thought about adding an organic phase and only injecting the top layer, but I personally think that not the entire sample may be transferred to the organic phase (partition coefficients, etc.).

Any thoughts or other suggestions...?


Thanks!

Roxanne.

[zokitano]

Dear Roxanne42,

If you are thinking about direct GC injection of the water samples containing those analytes I would suggest you not to do that.
Water compared to the other solvents have great vapor volume, which in turn can cause backflush of the sample in the injector gas lines if injected directly in GC injector (which is not desirable).
Besides that water can damage the stationary phases of the capillary columns especially when operating on high oven temperatures and using polyethylene glycol (PEG) phases (which are more susceptible on stationary film degradation or "column bleed").

I would personally suggest you to try to do headspace analysis of your samples. By this, you'll avoid injecting water samples directly on column and with proper optimization of the headspace conditions you can obtain the desired sensitivity for your analyses. Also you'll be able to use PEG stationary phase columns without the risk of column bleed (if you're working in the temperature limits given by the manufacturer, of course ). PEG phase would be suitable for separating 2-butanol from 1-chlorobutan-2-ol.

Hope this helps,

Regards

[Roxanne42]

Unfortunately, we are not equipped for headspace GC injections at the time (educational facility) also.

It may be something we will consider in the future but it is not feasible at this point.

Thank you for taking the time to reply though! Greatly appreciate it.

[oscarBAL]

Hello Roxane; I think something that could work is SPME, the graet adventage you can submarge your fibber directly in your solution; but the proble is probably you will need a little bit of training; the cost for hardware can reach $4K; so much cheaper than Head Space. and you maybe you can use SPME for others analysis.
Other alternative is use SPE; but you have to select you stationary phase (the same with SPME).

Letme know if SPME can work for you of you need mare information about it.

[kdamme]

It may be possible to do a direct injection of the water. We use Varian columns (CP-WAX and PoraPLOT) for the detection of 1-butanol, but it will probably also do well for 2-butanol and maybe for 1-chlorobutan-2-ol.

These columns are capable of having water as a solvent.

Detection limits are not too low. We detect +/- 2 mg/l on a FID. It might be enough for you? We inject 0.5 µl splitless on an old Chrompack-GC. More is not wise as zokitano said.

It's worth a call to your local Varian-dealer (or other column-supplier). And after they recommended a column, it might also be a good idea to contact your GC-supplier as we had problems with this application on some GC's.

[AICMM]

Roxanne42,

There is nothing that says headspace has to be done with expensive hardware. I have done and I know someone else who has done headspace by hand. All it takes is a vial and a syringe. Transfer the sample to a 20 mL VOA vial, let it sit for some time, use a gas sampling syringe and inject. Very effective and inexpensive. SPME can also be done in this manner for very little expense.

Besides, you have an autosampler. They're called students...... Just control the sampling protocol very well.

Best regards.

[s2008]

PLZ check the following reference and it might be helpful for you to consider direct aqueous analysis

1. food chemistry 75 (2001)101-108.
2. Analytical chimica acta 579 (2006) 88-94


Hope it will a little bit help

[Don_Hilton]

One direction you can take for analysis of alcohols in aqueous systems is the addtion of a salt to the mixture and then the use of extraction with a solvent or SPME. Solvent extraction and SPME both depend on partitioning of analytes between phases. The addition of a salt will make the water more ionic in nature and less "friendly" to the alcohols -- driving the alcohols into the SPME fiber or the extraction solvent.

[gcguy]

You can use a split/splitless injector, most columns I have seen can cope with a bit of water if they are bonded phases, water can actually be used as a solvent rinse. If the temperature of the injector is kept reasonably low then there should be no problem with flashback. There is a bit of free software on the Agilent website which will calculate vapour volumes at different temperatures and pressures.

Alternatively you could try an on-column injection then there would be no issues with flashback in the injection liner. You would just have to be careful of inorganics.
GCguy

[Peter Apps]

The strongest argument for using headpsace is the salts and reactants that are in the sample, which pose a more serious problem than the water. They will crud up the inlet in short order and, depending what they are, eat the phase off the column. Try the manual syringe headspace method, or manual SPME (you can even do manual SPME without the holder).

Peter

[Bruce Hamilton]

Depending on the experimental concentrations, I'd actually just try mixing with a large excess of a higher. immiscible alcohol ( or other immiscible solvent ), cool, centrifuge, and inject the solvent layer.

If the correct solvent and volumes are chosen, or you perform a couple of extractions and combine them, the differences in partition coefficients may be minimised, you can easily spike some samples to obtain a suitable calibration curve.

I'm guessing that the reaction concentrations are fairly high, so dilution is possible.

Please keep having fun,

Bruce Hamilton

[Roxanne42]

Thanks for all the replies!

After speaking with an Agilent rep - was told to go ahead and try water samples on our HP-1 column (stable to water). I was told there should be no problems as long as I watched the injection volumes.

Should the water remain in the vapour phase throughout the injection (i.e..; should the pressure and temperature be adjusted so)? We're using a split/splitless injector so we'll do split injections of 0.1 or so uL, with hopefully an inlet temperature no higher than 150,


Thanks,

Roxanne.