Chromatography Forum

please help. I have a problem with precision (% RSD > 3.0 %) of analyzing ethanol in water (solvent). my Area count of ethanol and n-propanol are up and down.

instrument : Shimadzu GC-2014 with AutoInjector

head pressure : 15 Psi initially hold for 0.5 min (I am using pressure control mode)
the pressure program is 15 psi hold for 0.5 min
then pressure down to 10 psi hold the rest of the run
inj. Vol: 0.5 uL
Split : 1:50
split Liner : Restek split liner
Column: ZB-WaxPlus (phenomenex, zebron)
30 m: 0.25 mm ID, 0.25 film thickness
temp: 50 C hold for 5 min
then ramp to 200 C at 25C/min

Ethanol RT : 4.6 min
n-propanol: 6.6 min

Double posting fragments discussions.

Peter

Yeah, but is the ratio of the ethanol to your internal standard varying, that's what is important, and it's the RSD of the RATIO that's important here (especially because of injections from water and its large vapor volume), not the rSD of the individual peal areas.

Thanks, my ratio is good.

Hi, I am still little confused about internal std calculation. do we need to consider the concentration of ISTD in the calculation. Also, for the RF, some literature use Area Analyte/Area ISTd. but some use Area ISTD/Area Analyte. IS it matter ? thanks

I would suggest stick with analyte/internal standard. In this way you see your data scaled relative to the internal standard. And a plot of area vs. internal standard for your calibration will look like an increase in response with increasing analyte concentration in the standards. Does it make a difference? It can in computing a linear regression because the magnitude of the error for specific data points changes based on which number is used as the divisor.

kka626,

Have you considered dilution with DMSO? I built an app for ethanol using DMSO to dilute the wine down. Does a number of advantageous things, puts analyte in more reasonable concentration range, drops the water on column way down, eliminates big split ratios, etc... Only disadvantage is the peak shape is not exceptional due to the fact that the ethanol elutes in front of the diluent. I still use an internal standard by the way, added to the DMSO up front so it is not an additional step in the prep.

Best regards,

AICMM

everything was fine . however, now I run it again. I have shoulder on ethanol peaks and internal (n-propanol) peaks. i changed new liner, new septa, cut 2 inches of the head of the column (injection side). But, still have shoulder on all peaks (ethanol and n-propanol). I might got contaminate somehow. please advice.

3% ethanol is a lot. Have you tried dilution with methanol? We find methanol elutes fully before ethanol on most polar columns we've used.

Thanks for response. No my %Rsd is less ok now less than 2%. I did the validation and very thing was fine. now, After a month, I run to test my sample, all peaks (ethanol & n-propanol) have big-belly shape, like waving. I change new liner, new sept, cut two inches inlet column. Same problem happen. (my new column recent use for my validation). Thanks. PM just perform on my system last week . Do u think detector problem ? Thanks

After a month, I run to test my sample, all peaks (ethanol & n-propanol) have big-belly shape, like waving. I change new liner, new sept, cut two inches inlet column. Same problem happen. (my new column recent use for my validation). Thanks. PM just perform on my system last week . Do u think detector problem ? Thanks

Not detector issue, if your gas flows are OK. I'd install a brand new column, and see if things are better, worse, or the same.

Most issues like this end up being column or inlet related.

Thanks, I will try that. Also, all peak heights are half compare to previous run.

Another column also show shoulder on ethanol and n-propanol. Changed inj. Port dwell time lower 0.3 sec (original 1sec) and higher 2 sec. Still problem.

Everything was fine. After the PM, then I have problem.

I injected standard manually, peaks come out nicely. Look like auto sampler generate this problem. I tried slow down the inj. Speed, still problem. Or will try inj dwell time then