which column is better for ethylene measurement?

[llonacho]

I want to study ethylene production and concentration in packaged produces (in MAP), so I have to measure ethylene in air. The levels I want to measure are about 0 - 5000 ppb, or more. I got two quotes, for two different columns, but I don't know which is better, or which I should choose. So I need your help with this .

1) Column Phenomenex 007-1 (http://es.scribd.com/doc/108045415)
2) Column Supelco Carboxen-1000 (http://es.scribd.com/doc/108045554)

I pretend to use FID as detector, and He as carrier (I can use Hydrogen or Nitrogen too).

Another question I would like to do is, is there a way to know the limit of detection? Because that information is not given in any tech data of both columns.

Any advice or comment will be well received

[chromatographer1]

llonacho

You have been given poor advice on both columns. I wonder if you gave either vendor the actual conditions of your testing or did you merely tell them "I want to measure ethylene in gas samples" ?

The thick film capillary column with a 50:1 split and 500 microliters sample injection will certainly have issues even seeing a 5000 ppb peak much less looking at lower levels. Then you can have variation of peak size discrimination issues which will occur whenever you reinstall the same or new column in your injector.

Likewise, using a 15 ft packed column and looking at 5000 ppb peak with a temperature programing of the oven will lose any peak height and you will only see a modulation of the baseline hiding the peak altogether.

You should use an A or D or N polymer (Poropak or Hayesep) that has been thoroughly cleaned before packing in 2 Fused Silica coated metal tubes of 1mm ID of 1 and 2 meters in length with installed screen or frit terminations in a two column backflush or heartcut configuration. (R or S are possibilities but are not recommended) This will give you a clean and flat baseline required for trace analysis.

A capillary column will require subambient temperatures to achieve good separation of the three C2 peaks.

You must separate ethylene from any acetylene that might be present, so the P or Q polymers are not good choices.

You should use a 0.25 - 1mL loop directly connected to your column train if possible with an isothermal oven, 40C to 80C temperature range. ( you decide which is best) I would also use nitrogen carrier (10-30 cc/min) if possible, although hydrogen or helium is acceptable.

If you choose to use a single column with no valves, then you will have to allow enough time at elevated oven temperatures to cook off the column thoroughly to remove the larger hydrocarbons from the column and allow enough time for the baseline to equilibrate before you execute another injection of sample.

good luck,

Rod

[Carvone]

An unripe banana maybe? With correlating the change from green to yellow

All joking aside, you could look up the references listed at the NIST Chemistry webbook for ethylene. There are a range of different columns used, the settings, and the Kovats Retention Index for each reference.

http://webbook.nist.gov/cgi/cbook.cgi?I ... #Gas-Chrom

[llonacho]

Thank you both, your answers have been very useful.

Rod, I have a question for you, are you sure there can be acetylene? because I read some articles (where the did what I want to do) and they used the Q polymer.
Which are the differences between the different polymers?

Suppose I want to make the test isothermal (just because it is simpler), and use Hydrogen (because I have the generator and is free, if it is not recommended I can use Nitrogen [cheaper than Helium]), which of the polymers would you recommend me?

[chromatographer1]

You should use an A or D or N polymer (Poropak or Hayesep) that has been thoroughly cleaned. You want only one in a recommendation?

Use Porapak N 80/100 mesh.

There is no loss in using a polymer that separates acetylene if price and performance are the same as Q. The question is not "Will acetylene be present", rather, the statement is "if it is present it will NOT interfere with an accurate measurement of ethylene". Acetylene and ethylene COELUTE on a Q or P polymer column.

The different polymer packings: A, B, C, D, N, P, Q, R, S, T, U, etc indicate that a different copolymer is used with styrene in the polymerization process, and sometimes a different pore chemical is used in manufacturing, thus the different temperature limits, pore volumes, and selectivity of separations.

If you use a single column (no Back Flush or Heart Cut valving) then use Hydrogen carrier, 30cc/min (no fuel will then be needed for your FID if you set your air flows correctly for 30cc/min hydrogen). You can of course, add 10 cc/min fuel and keep your standard air flow. If you do not use a backflush (NOT RECOMMENDED, a good backflush is a chromatographer's best friend), then your isothermal cycle time will be VERY LONG or you may have an unstable baseline if ANYTHING but ethylene is in the air. (hope that a truck does not drive by your lab, or that anyone is wearing aftershave in your lab)

If you use nitrogen carrier you will not see valving upsets in the baseline. If you use hydrogen carrier you will. This may not cause problems or it may.

Happy Research,

Rod

[llonacho]

A lot of thanks Rod, I'm taking notes. I'm trying to get a column Porapak N 80/100 mesh, and I will use Nitrogen.

You said both times that the polymer have to be thoroughly cleaned before packing, if I buy a new column I supposed it come clean?

I don't expect that large compounds to be present, but I suppose I will discover that in the trials, we don't have a backflush.

Thanks again, rod.

[chromatographer1]

When you order the column you should specify that the packing be cleaned before packing. I don't know if you can find a vendor that will do that for you. They will tell you that they sell the packing as Waters delivers it to them, and that will be a gamble.

And yet, usually, cleaning will remove contaminants that if allowed to remain and the column is heated, broader peaks will result as the contaminants will react within the pores and partially block them. Many vendors will have no idea what you are talking about. It takes a few days and a minimum of effort and cost to perform a good cleaning of the packing. I used to do it routinely. But I wanted the best possible performance. A 3 meter 1mm ID column should require about 1 gm of packing.

Also be aware that using glass wool plug terminations will increase the void dead space internally and will also reduce column performance. Hopefully you WILL purchase frits or screen terminations. A poor packing job will also develop packing voids after heating. Go to a quality vendor for your column.

Good luck,

Rod

[Johnny Rod]

J&W's GS-Alumina PLOT column is one we use to good effect on hydrocarbons up to C5 and above. To see 5ppm well you'd want 1ml injection and can use 10:1 split.

[llonacho]

Yes, I have seen more works using alumina than using porapack, I don't know which is better?.

For alumina is reported a 30m column, and for Porapak Q (80–100 mesh) a 90cm column. In this case I don't know how to choose the proper length in any case.

[gserrano_84]

All the suggestions that were provided are good. Just want to say that the difference in length is because the alumina PLOT is a capillary column, and the Porapak Q is a packed column. The PLOT columns use a layer of the adsorbent material, thus, doesn't generate as much pressure drop as the packed columns.

either will be useful, but as Rod said, with capillary columns you'll probably have to work at subambient temperatures for good separation of C2s.

Gustavo/Supelco

[chromatographer1]

And the difference in COST ?

Capillary - $450 ?

micropacked - $150

The alumina column is the best capillary choice.

And which will give GOOD and ACCURATE measurements at less than 5 ppm? or AT 100 ppb ?

best wishes,

Rod

[Johnny Rod]

You can add in helium/carrier costs if you want! Look up some J&W chromatograms, you can separate C1, C2 etc. at 40C or even above depending what else you expect to find. This we use for a UKAS-accredited method below 1ppm. Maybe with Plot-Q or similar you'd need subambient temps?

[llonacho]

Sorry, I'm lost now. Which is better, capillary or packed? (remember I need to measure levels about 5000 ppb or less [may be 20 ppb]). I don't have refrigerating system so my oven temperature must be above ambient temperature.

Johnny Rod, you said "UKAS-accredited method below 1ppm", which is the name or number of the method?, so I can search it.

Thanks

[AICMM]

llonacho,

I gave this a try just for fun. I used a packed n-octane on Res-Sil C as the pre-column feeding a packed Hayesep C in a backflush to vent configuration. I used argon as the carrier gas and the argon ionization detector I have developed as the detector. I was using a 2 mL loop and was shooting a 2.4 ppm standard with good response. The backflush configuration will eliminate the heavier hydrocarbons that may be present, while the C gives good separation of air, CO2 and ethylene. It should also resolve acetylene according to the RT tables I look at but I confess I cannot guarantee that.

Your sensitivity will be challenging without pre-concentration. Ethylene is only two carbon so FID response is not that great. Neither is AID for that matter but I wanted to demonstrate the ability to look at low level ethylene without any flammable gases (argon is non-flammable and relatively cheap.)

FYI, I was looking at jalapeno peppers (ethylene concentrations around 0.5 ppm) by headspace. My most striking observation was CO2 concentrations climbing after cutting up the pepper.

Best regards,

AICMM

[chromatographer1]

The problem with any capillary is sensitivity. Just because alumina works doesn't mean you get out what you put in. (alumina is reactive, that is why they try to deactivate it)

Micropacked porous polymer column will overcome this issue. You can use higher flow rates to enable larger sample loops (1-2ml) which enable lower detection limits. This means higher pressures however, which suggests hydrogen carrier is better.

Everything including price is a compromise. I have given you the best cost performance suggestion I can. You can always spend more and be disappointed.

best wishes,

Rod