Chromatography Forum

Hi,

Any suggestions to remove trace carbonyls/ketones from HPLC grade acetonitrile. I am getting both formlaldehyde and acetone.

I thought perhaps to add DNPH to the acetonitrile and try a low temperature distillation.

Any thoughts? The acetone is present as about .04 ug/mL and formaldehyde about .03 ug/mL.

In 1983, I used to prepare acetonitrile for trace carbonyl analyses by distilling it with a trace of DNPH and about a ml of 0.1M perchloric acid. It was very quick and relatively easy - and I used to purify about 2 litres in a morning using a simple Quickfit system of 3L flask, splash head to control aerosols, thermometer to discard low boiling fraction, and condenser.

I found that I didn't need any fractionating column as long as boiling rate was controlled - three antibumping chips added each time.

I was also able to recycle most of the mobile phase by azeotropic ddistillation, as acetonitrile was expensive here in those days, but can't recall the details.

However, by the late 1980s the available HPLC grades were so low in carbonyls, that I didn't need to process them.

My first suggestion would be to check to ensure that you have fresh acetonitrile and are avoiding any volatile contamination, and are purchasing the best possible grade for carbonyl analysis, and especially try products from different manufacturers.

I found some manufacturers' HPLC acetonitrile grades were very low in carbonyls, whereas some others were quite high, even in their best grades. These days, with very high purity, low UV grades, they should be low in carbonyls.

There were lots of paper published on atmospheric and emissions carbonyl analysis that included details of carbonyl removal from solvents were described.

Please keep having fun,

Bruce Hamilton

It is also important to realize that formaldehyde is ubiquitous in the environment now. There are many sources (clothing, carpets, certain wood products, etc.), so it is possible that some of the formaldehyde is coming from the air. This would affect your water as well; perhaps even more so. In some labs it may not be possible to get a "zero" blank.

So, even if you can remove the carbonyls at some point, you must be careful to protect the solvents from further contamination.

Good luck.

Thanks for the replies.

To Bruce regarding specifics for the process he used in 1983:

The recipe:
1) 2L acetonitrile
2) 1 mL of 0.1 M Pechloric (in water)
3) "Trace of DNPH" --
not sure how much this is.

Assuming:
For 2L of acetonitrile containing approximately
.04 mg/L formaldehyde (30.03 g/mole),
0.03 mg/L acetaldehyde (44.05 g/mole),
0.1 mg/L acetone (58.09 g/mole)

the total moles would be:
formaldehyde: .003 Moles
acetaldehyde: .0007 Moles
acetone: .002 Moles

Since the reaction with DNPH is 1:1, and to double or triple the available DNPH, I'm thinking to add about 0.015 moles of DNPH
(at 198.14 g/mole this would be about 3 grams in the 2L of solvent)

The apparatus:
Not sure what a "splash head to control aresols" is.

Comments?

As noted earlier, the easiest solution would be to purchase a low UV HPLC grade from a suitable supplier, as I expect that several brands will have quite low carbonyls. That would be my first choice these days.

Your calculations are out by 1000, think you require a minimum of about 1.5 mg, but I would add about 10 times that to cover other consumption. I would have added about a pinch ( 20 - 50 mg ) of DNPH, but I may have gone up to about 100 mg of DNPH for some acetonitriles. You can probably add less perchloric, but that quantity worked for me.

You can search on Google for images of the Quickfit items, and there will be equivalents from other suppliers. The Quickfit system ( catalogue numbers ) that I used would have probably comprised of:-

Three litre single short necked flask with 24/29 socket
Splash Head - Vertical ( SH 17/23 )
Stillhead, Plain ( SH 4/22 )
Thermometer Adapter ( ST 51/13 ) with a suitable thermometer
Condenser, Liebig, long ( C1/22/SC )
Adapter, Receiver ( RA 3/23)
Two liter flask with 24/29 socket

I'd normally use 3 anti-bumping granules for each distillation. They can't be reused, and should be added to the solvent just before heating is started. Never add anti-bumping granules or any fine solids to a hot solvent.

You should discard the first 5 - 10%, or until the thermometer is stable, and leave the last 10 - 20%, or whenever the thermometer has risen by about 0.5C.

Note that it's faily important to have an efficient condenser and minimise air access to the distillate, which may be warm. I did use a small stream of nitrogen via a 1/16"OD PTFE line fed into the RA3/23 vent sometimes ( obviously leaving space around the line for the gas to escape ), but found no difference in carbonyls, and often didn't bother.

I'd be surprised if it is ambient contamination, as acetaldehyde is also high. It's sensible to take precautions though, as any NOx from nearby combustion sources will introduce a peak just before the HCHO-hydrazone.

Please keep having fun,

Bruce Hamilton
“Aim high, there is little virtue in easy victoryâ€