By Anonymous on Thursday, June 10, 2004 - 06:48 am:

Dear All:

I am an undergraduate student trying to develop a method for ethanol analysis, mostly at high ethanol concentrations. I am using HP 5890 Series II with FID.
Column: HP-5 (5 % Ph Me Silicone) 10m*0.53mm*2.65um film thickness.
I prepare varying ethanol concentrations from 80 to 95 wt%) and then I make a sample prep in a small vial such that the ethanol: water is in the ratio of (1:10). Then I inject the sample manually. I am having serious trouble in getting a good fit for my data. The regression value ranges between 0.8 to 0.9. I have developed a graph with mass fractions on the x axis and total area on the y axis.
I am not using any internal standard.

Any suggestions and help will be welcome. Please post your expert advice in trying to develop a method for ethanol.

Thanks

-------------------------------------------------------------------------------------------------------
By Anonymous on Thursday, June 10, 2004 - 08:16 am:

I am not suprised your fit is not good.You may meed a more polar column and use n-propanol as an internal standard - such as a wax type column.
Have you access to an autosampler? That would help too.

GC Dave

-------------------------------------------------------------------------------------------------------
By PPG on Thursday, June 10, 2004 - 08:17 am:

In FID ethanol give a very bad result. The only thing is to do test samples and curves if you are every time with the same products around.

ppg

-------------------------------------------------------------------------------------------------------
By Anonymous on Thursday, June 10, 2004 - 08:33 am:

Not to sound too critical but I would bet your problems would all but disappear if you used volumetric glassware and an analytical balance. Trying to do quant work in little vials is just asking for trouble. The suggestion from above to use an internal standard is good, helps with the injection volume variations. If you have good peak shape right now, there really isn't the need for a different column, I have analyzed mixtures of alcohols on HP-1 and HP-5 columns successfully.
Regards,
Mark

-------------------------------------------------------------------------------------------------------
By Consumer Products Guy on Thursday, June 10, 2004 - 09:40 am:

We use polar capillary columns (PEG-type, 10m x 0.53, with autosampler) and n-propanol as internal standard, similar to that mentioned by first responder. Keep injection volume low (0.5 ul) because water expands so much in the inlet. We have used the same vials on HP-1 columns as suggested by other responder, but peak shapes are not as good.

-------------------------------------------------------------------------------------------------------
By Anonymous on Thursday, June 10, 2004 - 10:18 am:

Dear All:

Thank you all so much for the suggestions.
I have in the past used n-propanol as the internal standard. The column length being 10m, showed very little separation between ethanol and n-propanol. So, I tried changing the internal standard to many other alochols and went upto pentanol, but still when I try to get a response factor for just the internal standard i.e., ( x-axis: varying concentrations of internal standard, y-axis: total area count)the R2 value is less than 0.9. So, i decided not to use an internal standard. But still I will go ahead try once more with n-propanol as the internal standad.
Can you please suggest a suitable concentration for n-propanol? In the past I had used 50% of n-propanol.
Also, I use an analytical balance of very high precision (4-place).
The operating conditions I use are:
*Initial oven temp = 100
* Rate = 10 deg/min
*Initial hold up: 0min
* Final temp: 150 and final time: 1min
* ZER0 = 20
* Range = 10
* ATTN = 7

Please let me know if these settings indicate any potential reasons for my problems.

Thanks once again

-------------------------------------------------------------------------------------------------------
By Anonymous on Thursday, June 10, 2004 - 10:57 am:

Dilute your sample to around 0.1% of ethanol, and set split ratio to 1:50. Concentration of propanol must be as that of ethanol.

-------------------------------------------------------------------------------------------------------
By Anonymous on Thursday, June 10, 2004 - 11:20 am:

Try these conditions:
Flow 5 mL/min
Injector: 70º C
Detector: 260º C
Oven: 35º C for 5 min, then increase to 175º at 8º per minute (you may want to put a burnout time in)
Inject 1 µL at a split ratio of 1:50

Al

-------------------------------------------------------------------------------------------------------
By Anonymous on Thursday, June 10, 2004 - 11:33 am:

Dear All:

Thanks for the suggestions.
I have another question...I weigh accurately to 4 places the mass of ethanol mass fraction (eg.90%)and add water equivalent to 1/10th of the mass of ethanol taken in a sample vial. If I need to add an internal standard such as n-propanol, i intend to use 90% concentration of n-propanol (since i am trying to analyze ethanol at very high concentrations). What should be the mass of n-propanol that I should add in the sample prep (vial). Should it be the same mass as that of the the ethanol or what should be the ratio of it to that of the ethanol?

Thank you very much...

-------------------------------------------------------------------------------------------------------
By Anonymous on Thursday, June 10, 2004 - 11:36 am:

While you are not using analog output never mind zero, range and ATTN.

Anonymous on Thursday, June 10, 2004 - 10:57 am

-------------------------------------------------------------------------------------------------------
By Anonymous on Thursday, July 1, 2004 - 04:45 am:

I agree with anonymous, that diluting is possibly better than changing the split (the expensive dilution).
50:1 is a common split ratio

-------------------------------------------------------------------------------------------------------
By Anonymous on Thursday, July 1, 2004 - 12:23 pm:

What is your injector temp?
What is your type of liner?

If you try 5 injections of the same Standard,are they reproductible?

If no You might have injection problem...
Autosampler would help and a istd too,or liner packed with glass wool,laminar liner...

-------------------------------------------------------------------------------------------------------
By Anonymous on Thursday, July 1, 2004 - 01:37 pm:

With a 10 m megabore column it will be difficult to run a high split without increased the column head pressure to a value which gives a very high flow rate. For the conditions listed above a 2 psi head pressure give a linear velocity of 53 cm/sec, which is getting to the high end of the range you want to be in. For the suggested conditions above the pressure would be approximately 1 psi, which most controllers will not hold that accurately.

Using the conditions suggested above with a 30 m column should work fine. You might ba able to get by with a 10 m column using the wax type column suggested above, but a longer column will be easier to develop the method on.