Very tiny peaks

[Jumpshooter]

I am trying to separate and quantify a series of 5 chlorophenols on a GC, injector temp = 250C, oven ramp, He flow = 2.2 ml/min; 40C 10C/min to 320C, Detector 350C, column HP-5 (5% diphenyl 95% methpolysilox.). The stock STDs were made up to be 1% w/w in the solution of IPA,di upon which I diluted them 1:100 with IPA,di prior to injecting. At that conc'n I should be seeing fairly large peak areas; however, I only see peak areas in the range of 25 to 45 mAU. I have to really raise the attentuation to even view them for the purposes of computing Tailing and Rs. Clueless as to why such large amounts of chlorophenol injected on-column are giving such a small peak response?

[willnatalie]

what kind of detector are you using to anaylze these compounds?
"At that conc'n I should be seeing fairly large peak areas" Where are you gettting this info? based on previous work.. manuscripts you've read?


Will

[Jumpshooter]

flame ionization detector FID in use here.
Well, my reference is the product literature that came with the HP 5 column--which I believe they have no reason to fabricate?

[willnatalie]

that second point was just out of curiosity.
Is the jet dirt?
Manual of Auto injection, either way is there any problems with the syringe?

Just some quick Ideas.

hope this helps.

[Jumpshooter]

None of the posted explanations provide a suitable explanation of the observed phenomenon.
The syringe is clean, the injector is not fouled. Everything else works fine on the analysis of other analytes.

[Don_Hilton]

1) Have you injected something other than the phenol solutions into the instrument to see how the instrument responds to other compunds?

2) have you used a leak detector to check all the connections from inlet to the detector?

[Peter Apps]

Assuming that the dilutions were accurate you inject 100 ng per microlitre into the inlet. If you inject 1 ul and this all gets onto the column you can expect good sized peaks. How much do you inject ? How is the inlet set up, type of liner and packing (if any) and when was it last cleaned / exchanged ? Split or splitless, and if split what is the split ratio, if splitless at what time does the split open ?.

How are the peak shapes ? tailed ?, overloaded ?

I am puzzled - mAU is milliabsorbance units, which is appropriate for optical detectors as used in HPLC, but not for FIDs, and it is a measure of peak height, not area. Increasing the attenuation makes peaks smaller, not larger.

Peter

[Don_Hilton]

Now that I completely read the post -- you say that the peaks are fine for other analytes. I assume that these are other analytes in the same mixture? If so, I would remake the standard and check again. This problem sounds familiar - it was either a wrong weight or a wrong dilution in the case I remember...

[AICMM]

Jimpshooter,

I'm with Peter, I want to know splitless or split and wool or no wool. At 20:1 split you're only putting 5ng on column which is really not all that much for an FID, especially for reactive components.

Best regards.

[Jumpshooter]

I truly appreciate all the fine comment to our puzzling issue here. Let me specify:
The phenols STD is a mixture of 6 chlorophenols. The working STD is prepared gravimetrically dissolved in a diluent comprised of 50% IPA/di. I weigh out 0.10 g of each crystalline STD into a small tared amber vial and QS the final weight sol'n to 10.00 g with the diluent (~9.395 g of diluent was used). Theoretically this gives me a 1% w/w sol'n of eac phenolic--yes?
My split ratio is set at 20:1, so I'm getting about 50 ng on-column with each injection. The run parameters were previously cited.
The column may be too non-polar for these slightly polar compounds?
Saturday I injected each single analyte and observed very small peaks also ( will try to post images). The peaks did not appear corrupted by either tailing or fronting--but they were so so small.
Is the problem with our FID unit?

[Peter Apps]

Your first post has you diluting the 1% solution by a further factor of 100. Are you actually doing this, or do you inject the 1% solution that you make up gravimetrically ?

By my calculation if you inject the 1% gravimetric solution with a split of 20:1 you have 500 ng of each on column. If you inject a 1 in 100 dilution of that 1% you have 5 ng on column, which will give you a decent sized peak, but will be way more susceptible to adsorption on active sites.

Just to be sure - do you put all six phenols into one vial and then add the solvent ?, or do you make up 6 separate 1% solutions and them mix them ?

If you make an equivalent dilution with something nice and tractable like an alkane or an ester, how big is the peak ?

500 ng of a phenol will give a front-tailed peak.

Peter

[Peter Apps]

Your first post has you diluting the 1% solution by a further factor of 100. Are you actually doing this, or do you inject the 1% solution that you make up gravimetrically ?

By my calculation if you inject the 1% gravimetric solution with a split of 20:1 you have 500 ng of each on column. If you inject a 1 in 100 dilution of that 1% you have 5 ng on column, which will give you a decent sized peak, but will be way more susceptible to adsorption on active sites. 500 ng of a phenol will give a front-tailed peak.

Just to be sure - do you put all six phenols into one vial and then add the solvent ?, or do you make up 6 separate 1% solutions and them mix them ?

If you make an equivalent dilution with something nice and tractable like an alkane or an ester, how big is the peak ?

Peter

[Jumpshooter]

After perusing all of your very kind posts I have gone back and re-done my entire work-up from scratch. I prepared freshly weighed STD's (per the comments on possible math or weight errors) and I have checked my run parameters (per comments on that factor). I got the same results, then I changed columns to another of the same type (HP-5)that we have in our inventory. I used the same type because none of the posts indicated that the column phase was the source of the problem. When I re-ran the samples I got larger peaks more in line with what you'd expect if >50 ng of analyte is presented to the column. I concluded that the other HP-5 column was bad/defective or had become corrupted somehow.
I do remember that when I pulled the previous column out of the box it was stored in that it did not have the rubber septum endcaps on it. I guess these columns are more facile than I once thought?

[AICMM]

jumpshooter,

What is the density of the final mixture you are shooting? I assume you are shooting 1 uL (or so) with that 20:1 split so Peter is right in that you are really only putting about 5 ng on column. For an FID, I would argue that is not all that much.

I have seen columns eat chlorophenols in the past so that is certainly a possibility.

Best regards.

[Jumpshooter]

AICMM: OK, it appears that some re-consideration of called for in the sample prep.
1.001 g of clorophenol was brought to 10.002g final sol'n weight with 50%IPA/di. The density of the diluent was 0.882 g/mL; so 10.20mL of diluent was used. This made the phenolic sol'n conc = 1.001g / 10.20 mL =0.098g/mL in clorophenol. I tasked the GC to inject 1uL with a split ratio of 20:1, so the amount of phenol on-column = (1uL)(0.098 g/1mL)(1mL/1000uL)(0.05)=4.9ug on-column. This amount is more than enough to give a decent sized peak no?