Ethanol by GC

[mahmoud]

aa Hi all,
I am analyzing ethanol (1000 ppm) in a solution (acidified water and acetonitrile). Ethanol peak separates very well from other components but the problem is the peak shape and reproducibility of injections.

Thanks
Mahmoud


The parameters used are as follows (I tried head-space before with %RSD of 20 for 6 injections):
Chromatographic Conditions
Column: 75 m x 0.53 mm I.D., 3 µm thickness, fused silica column coated with Rtx-624 stationary phase or equivalent
Column Temperature: Initial: 40°C -hold 0 minutes
Ramp 10°C /minute to 240°C -hold for 0 minutes
Injection Temperature: 180°C
Detector Temperature: 250°C
Carrier Gas: Helium
Carrier Gas Flow Rate: about 5.0 mL/min
Injector Split Ratio: 4:1 tried also 20:1


LEAP Conditions
Cycle: GC-Inj
Syringe: 10 µL
Spl vol: 10.0µL
Air vol: 0.5 µL
Fill vol: 5 µL
Fill speed: 2 µL
Fill Stroke: 5
Pull-up: 5 s
Inject to: GC-Inj-1
Inj. Speed: 5 µL/s

[chromatographer1]

Mahmoud

Thanks for great details of your method. Some critical items were missing:

the details of the vial you are using, volume, material, septum type

temperature and time you are heating the vial

If you are using 10 microliters of sample, that is small but not too small.

I would @ 80°C for 10-15 minutes.

1000 ppm is about 1µg/µL of ethanol which means you are putting about 10µg in the vial. That is a lot for a small vial and reasonable for a 22mL vial.

RSD values of less than 3% should be possible if the vial and needle are heated properly. Reproducibility is determined by how well your temperature control is maintained.

I have never used a LEAP, but PE, Varian, or Tekmar analyzer only, so I hope my advice is sound.

best wishes,

Rod

[mahmoud]

Thanks Rod for the reply.
I am using a 1 µL injection which is about 1µg on column and am splitting, too. Could it be an inlet or detector problem? I used headspace (headspace vial 20 ml, filled with 5 ml solution) and direct injection using standard glass vials, whcih are similar to those used in HPLC and capped firmly. I have changed the thermal septa but with no luck. I tried a number of incubation times and temperatures, also I tried different needle temperatures.

Thank you

[Chandresh K. Soni]

I think if u starts analysis From 50°C and Ramp @ 20°C then The Peak shape of EtOH is Sharp.
U can use also Non Polar Column Like CPSIL-5CB , Ultra-2
With 0.25mm I.D. and 0.25µm Film Thickness Column u Can get More Sharpness.

[AICMM]

mahmoud,

If I understand you correctly, you have tried both direct injections and headspace injections? Regarding direct injections, try diluting your sample in 1/2 with isopropanol. A trick someone I know tried which greatly improved the flash and therefore the reproducability. Will elute after the ethanol on the DB-624. Also, make sure you have a large bore liner and probably better to use the 20:1 split.

Regarding headspace. You have 5 mL of a 1000 ug/ml solution in a 20 mL vial, right? So, assuming you have 100% vapor phase partition (a very poor assumption for ethanol) means you have 5000 ug/15mL of headspace or 333 ug/ml or 333ng/uL of headspace. So at 1 uL or headspace and 20:1 split, you are only putting 16 ng on column. If I understand your post correctly.

Best regards.

[chromatographer1]

Mahmoud,

Things are a bit confused for me.

Let's work on 1µL liquid sample injection first:

First comment, a 75 meter column is much more than needed. But ok.

If injecting 1µL water solution, don't inject that much. reduce your injection volume to 0.3µL or less.

Use a Chaney adapter if possible.

High RDS values may be due to vapor flash overfilling the injection port volume. Solution inject less.

Don't shoot acids on capillary column unless they are organic acids, not mineral acids. Even then think about it.

Headspace: Manual or automated?

Manual ? Don't. Think about it again and find another way. Some have no problems with it. Most do.

Automated ? It is wonderful. Use 250µL loops or syringes with good thermal control of vial and syringe heating, timing, and injection.

Alternative? Packed column with autosampler. RSDs will be less than 2%, often less than 1%.

Simple easy and quick. 10ft Porapak Q 80/100 mesh used for years, still used, why? easy, quick, difficult to make fail or lose accuracy.

CHEAP

best wishes,

Rod