SPME-GC-MS problems, help

[ableich]

To anyone with experience using SPME-GC-MS,

I have a started to use SPME to analyze aromatic Gasoline Range Organics (BTEX,) and MTBE in Groundwater, however, I'm running into problems. Reason for picking SPME are obvious fast, cheaper, not as complicated as purge and trap and results are comparable to P&T. The fiber I'm using is a supelco 85 um Carbowaxen/PDMS, which was used in a paper that I'm following published in the: Journal of Chromatography A, Volume 1021, Issues 1-2, 22 December 2003, Pages 157-164
Fu Fang, Chia-Swee Hong, Shaogang Chu, Wenpeng Kou, Anthony Bucciferro.
The GC injector is equipped with the 0.75 mm ID inlet liner and I'm using the 11mm LB-2 septa. The capillary is Supelco VOCOL (10m x 0.2 mm x 1.2 um). The problem that I'm running into is that I'm getting severe band broadening and tailing for most of my peaks (some peaks even appear to have "Christmas Tree" shape), and I'm also having problems extracting MTBE. This has occurred with my 250 ug/L standard. I conditioned the fiber according Supelco. I believe I'm having problems focusing the analytes onto the column. The conditions are the following:
SPME-
1.Place 10mL of sample in 15 mL SPME Vial.
2. Add NaCl and stir bar, incubate and stir at 60 C
3. Insert SPME man holder and press down to allow fiber to absorb HS, extract for 10 min
4. Desorb in GC injection port for 2 min at 220 C

GC/MS
1. 35C for .5 min, ramp to 200 C at 25C/min hold at 200C for 1 min.

Before moving any forward, I descided to inject 1 uL of my PDS in the instrument with a split ratio of 100:1. All the peaks appeared sharp, as they should be, and are eluting about 0.5 min later. To make matters even worse is that I'm doing this all on brand new Agilent 5975c GC-MS, and I'm under a deadline to get this method developed and validated. Could the fiber be damaged already? Should I try starting at 50 ug/L?

[danicrd]

Hi
what are your gc conditions? Flow, split, inj tem, etc?
This problem seems to me flow related: maybe you have not flow enough to desorb in a tight band your analytes.

[Suresh Seethapathy]

Hi danicrd,

Are you using a precolumn? Are you using the splitless mode and what is your splitless time? You have a pretty thin stationary phase thickness which might not necessarily be enough to focus (i might be wrong). Reducing the concentration will give you an idea about that (or expose the fiber for a very short time with the same concentration).

Good luck
Suresh.

[Bigbear]

When we use SPME in small diameter columns we use cryo focusing with a JASCO trap ( installed in the oven and column goes thru it). I think you will get band broadning without cryo focusing.

good luck
Bear

[AICMM]

Ableich,

You are using 250 ng/ml standard at 10 mL. That is 2500 ng of analyte. If you are any where near 100% efficient you are putting a lot of material on column, way too much for the film you are using. Why not try 1 or 2 mL of this same standard in the same vial? Even then, the column may have some band broadening (although I can understand why you chose this column) so a split injection might also be in order here.

Best regards.

[chromatographer1]

There are several issues that may cause the problem that is occurring. It must be noted that Carboxen-PDMS fiber is an adsorbent type fiber that traps analytes in pores and retains them until thermal desorption. The desorption process with adsorption fibers is not as efficient as with absorption type fibers such as PDMS or PEG. Adsorption type fibers are better for trace concentrations because there is a limited fiber capacity based upon the size of the analytes. Here are my suggestions

1. The desorption temperature is too low, it should be at a minimum of 280°C. It will work even better at 300°C. Don't worry about the column. The column only goes several mm into the injection port and the bottom of the injection port is the coolest part. The column will not be damaged. Changing this parameter alone will probably correct your problem
2. The concentration of the sample is high for this fiber, especially for nonpolar analytes. You might consider using a lower concentration for this application. The use of salt may not be necessary but it will enhance the extraction of MTBE
3. Open the split vent after 0.75min. This will reduce tailing and improve peak shape. The other option is to use a split injection set at 10:1. This will increase the linear velocity through the injection port and sharpen the peaks. You may actually get better sensitivity because of the improved efficiency. When instrument manufactures started to split purge trap samples, the lower detection limits improved because of the higher flow rate through the trap leading to better efficiency.
4. Reduce your extraction temperature. For this application 45°-50°c is ample.
5. How is the stock standard being spiked into the sample? If the stock standard is prepared in methanol, you want the stock standard to be fairly concentrated so that the amount of methanol spiked into the standard is not too great. Methanol will have little effect on the extraction BTEX, but MTBE which is more water soluble will be affected by the amount of methanol. For a 10mL sample try to reduce the amount of methanol to 5µl or less.

hope this is useful
Bob Shirey (Supelco)

[ableich]

Thank you all for you sugestions. I might also not that a part of fiber broke in the injection port liner. Thus making it seem that I'm "twice" instead of once. I will change the inlect temp to 300C and change the split to 10:1. I'll let everyone know how I made out.

Thanks,
Alex