Separation alcohols

[chromos]

Hi all,
Does anyone please have any suggestions on how to separate ~95% ethanol and 2-3% C3-C5 alcohols and 2-3% gasoline on a system with precolumn (db-1 30m,0.53mm,1µm) followed by a polar column (cp-lowox 10m,0.53mm,10µm) where oxygenates are introduced to the lowox and gasoline components are backflushed through the precolumn. No dilution would result in overloading and perhaps overlapping with ethanol(?), and diluting would reduce signal from other components almost to noise(?).

[AICMM]

chromos,

This is a classic ASTM D5501 analysis so I am wondering why you are doing it this way. Is this a system you already have set up and you are being forced to use or what?

Beyond that, a couple of comments. Your DB-1 is probably not thick enough film to really hold up the light hydrocarbons but if they are only 2-3 percent it may not matter. Light hydrocarbons come out very early on the LowOx. What valve have you got installed, 4, 6, 8 or 10 port? Do you have a second detector available (makes life much easier)? Do you care about measuring the hydrocarbons?

Finally, you are likely to see some pretty significant retention time shifts with that much ethanol.

Best regards.

[chromos]

Thanks AICMM,

Actually we use this setup for analysis of MTBE in gasoline (ASTM D4815), and I am investigating the possibility to perform the mentioned analysis without any hardware modifications. A Deans switch is used to backflush heavier gasoline components to vent, and yes, as you said, light hydrocarbons come out very early throught the LowOx.

[AICMM]

chromos,

So you want to measure the ethanol or the impurities? You are trashing the heavy impurities on the backflush, right? So they can't come into the calculation because you don't measure them. You can area sum the light hydrocarbons since they will come out very early in the chromatogram so you at least have a handle on those. The alcohols will elute well on the LowOx but I would be concerned about retention time shifts of these components due to the high level of ethanol. I would especially worry about the components that elute before the ethanol. The ethanol peak shape will also be very poor since the LowOx does not have much capacity (not it's intended role....)

On a practical note, I would suggest as high a split as you can get away with and still see the levels you are looking for. The FID is reasonably sensitive so 1-2% levels means a pretty decent split ratio. I would also suggest you think about a heartcut configuration instead of a instead of the traditional 4815 configuration if that is possible.

I am working on a similar configuration (DB-1 and LowOx) except for low level oxygenates in hydrocarbon streams (alcohols in crude) so I will be very interested to hear about your experiences.

Best regards.

[larkl]

We use a similar setup for oxygenates in hydrocarbon samples. We use a very short "stripper" column 5m. I think 30m at 1 um should work fine. One thing to be aware of that is that the first oxygenate (acetaldehyde) will elute past some carbon number (C15 - can't remember), which would lead one to think that perhaps you can get by without a stripper column. Problem is the unsaturates/aromatics (ie - more polar) constituents will elute later into the oxygenates on the Lowox. We found spiking a normal sample with acetaldehyde only and then playing with the stripper valve timing to be critical to ensure all the hydrocarbons are removed. Then some time spent on the temperature programming the Lowox. You're going to have problems with the high level of ethanol. Seems like the chromatography is kind of "busy' on the Lowox around the ethanol. Going to be a bear to get the separation.

[chromos]

Thanks guys,

AICMM, it's mainly the ethanol and other C3-C5 alcohols that are of interest, while hydrocarbons are not (of course they are quantified as well if the alcohols are). Do you have any suggestions on heartcut configurations?

Best Regards

[cody84]

We use a ZB-624 for alcohols, no problems at all. 79% IPA or Alcohol.

[WK]

Chromos,
I agree a 624 column is excellent for low MW alcohols eg ethanol methanol ipa and n-propanol etc.
WK