-
- Posts: 45
- Joined: Sat Oct 12, 2013 9:58 am
Hello,
I encountered a strange problem. Two weeks ago everything was fine with the BTEX method, and now suddenly I get a huge decrease in peak height for the same amount.
The STD and ISTD solutions were prepared 6 months ago with a concentration of 10 mg/l. They are prepared in acetone to be able to spike water but also prepare injection standards in pentane. I experimented with 700 ug/l (700 ppb). Two weeks ago the Full Scan spectrum looked like the top row, and the 91 ion looked like the second row. Now it looks like 3rd and the 91 ion XIC like the bottom one.
The ratio is fixed so first thing is to notice is how little the bottom peaks are for Ethylbenzene and Xylens. Second thing to notice is the huge horrible peak at 9.01. Remember, this is the SAME standard solution I injected 2 weeks ago. Since is full scan I did a database search and identified the 9.01 to be:
The small peak at 9.27 seems to be:
Well, the original CRM ampule had the compounds in DMSO. The 2-Pentanone, 4-hydrohy-4-methyl I really don't know how it got there. Maybe is secondary reaction of the pentane with acetone or so...
My first reaction was to check all the ferrules, replace septa and liner, recondition the column. Nothing changed. Second thought was the ion source got dirty since I injected some horrible PAH samples and even with the backflush ON, some contaminants got through. But this does not explain the DMSO peak. Also I ran a PAH QC and for PAHs everything is fine sensitivity wise. Same goes for chloroform or DEHP. The only problem lies with the BTEX.
Do you think is possible, if some lab assistant left the standard solutions outside the fridge for a day, to destroy them ? I mean, it looks like the acetone vaporized and the DMSO concentrated while the BTEX vaporized too.
This is how it originally looked 2 weeks ago in SIM, the order is: Benzene, Toluene-D8, Toluene, Ethylbenzene, m,p-Xylene, o-Xylene.
Now it looks like this:
Sensitivity is down one order of magnitude, and the toluene peaks are horrible, with a huge tail.
To be sure the column is still OK, I prepared a 700 ppb BTEX in methanol (new CRM) and did a manual headspace analysis, 10 ml water, 30 min equilibrium time, 1ml injection, same GC conditions like for liquid injection. Resulted this full scan:
So the column is OK.
For fun I injected the 700 ppb in menthanol. Resulted this in SIM:
Conclusion: never use methanol as a GC solvent
So, what might be the problem with the BTEX in pentane ? Why the sudden sensitivity drop and the horrid peak shapes ?
Is the standard solution compromised or could be still a device loss of sensitivity ?
Notes:
- I use a backflush system with 2m deactivated non coated precolumn.
- I don't have automated headspace autosampler, this is why I use the liquid-liquid method also for BTEX.
- I use a 30 m DB5ms type of column (some Thermo equivalent) 0.25 um coating.
- For this BTEX method (both LL and HS) I use a 50:1 split ratio
Best regards,
Vlad
I encountered a strange problem. Two weeks ago everything was fine with the BTEX method, and now suddenly I get a huge decrease in peak height for the same amount.
The STD and ISTD solutions were prepared 6 months ago with a concentration of 10 mg/l. They are prepared in acetone to be able to spike water but also prepare injection standards in pentane. I experimented with 700 ug/l (700 ppb). Two weeks ago the Full Scan spectrum looked like the top row, and the 91 ion looked like the second row. Now it looks like 3rd and the 91 ion XIC like the bottom one.
The ratio is fixed so first thing is to notice is how little the bottom peaks are for Ethylbenzene and Xylens. Second thing to notice is the huge horrible peak at 9.01. Remember, this is the SAME standard solution I injected 2 weeks ago. Since is full scan I did a database search and identified the 9.01 to be:
The small peak at 9.27 seems to be:
Well, the original CRM ampule had the compounds in DMSO. The 2-Pentanone, 4-hydrohy-4-methyl I really don't know how it got there. Maybe is secondary reaction of the pentane with acetone or so...
My first reaction was to check all the ferrules, replace septa and liner, recondition the column. Nothing changed. Second thought was the ion source got dirty since I injected some horrible PAH samples and even with the backflush ON, some contaminants got through. But this does not explain the DMSO peak. Also I ran a PAH QC and for PAHs everything is fine sensitivity wise. Same goes for chloroform or DEHP. The only problem lies with the BTEX.
Do you think is possible, if some lab assistant left the standard solutions outside the fridge for a day, to destroy them ? I mean, it looks like the acetone vaporized and the DMSO concentrated while the BTEX vaporized too.
This is how it originally looked 2 weeks ago in SIM, the order is: Benzene, Toluene-D8, Toluene, Ethylbenzene, m,p-Xylene, o-Xylene.
Now it looks like this:
Sensitivity is down one order of magnitude, and the toluene peaks are horrible, with a huge tail.
To be sure the column is still OK, I prepared a 700 ppb BTEX in methanol (new CRM) and did a manual headspace analysis, 10 ml water, 30 min equilibrium time, 1ml injection, same GC conditions like for liquid injection. Resulted this full scan:
So the column is OK.
For fun I injected the 700 ppb in menthanol. Resulted this in SIM:
Conclusion: never use methanol as a GC solvent
So, what might be the problem with the BTEX in pentane ? Why the sudden sensitivity drop and the horrid peak shapes ?
Is the standard solution compromised or could be still a device loss of sensitivity ?
Notes:
- I use a backflush system with 2m deactivated non coated precolumn.
- I don't have automated headspace autosampler, this is why I use the liquid-liquid method also for BTEX.
- I use a 30 m DB5ms type of column (some Thermo equivalent) 0.25 um coating.
- For this BTEX method (both LL and HS) I use a 50:1 split ratio
Best regards,
Vlad
