help needed with gcms detection limits

[buggygcms]

Hello everyone.

I have inherited a 6890/5973 gcms that was used for VOC analyses and solvent extracts and analytical standards. I did not know the full history of the machine, but I did know that it had been awhile since it had been serviced. So I changed the column, liner, and septum.

So I went to run some hydrocarbon standards that were 100ng, but I am having problems. The peak of one compound got up to 18000 in abundance, but the baseline is very noisy and looks bad (similar to this: https://www.flickr.com/photos/129144858 ... 107490714/ )

I need my instrument to be able to detect <100ng (with a clean baseline ideally) and I'm not sure what to do.

Here are other details:
Column: db-5ms 30m, 0.25mm, 0.25um
inlet pressure:8.1
purge :10.0
toal:14.3
column flow:1.1


Autotune report from today:

Mass: 69 219 502
Ab: 386107 404042 55367
PW50: 0.60 0.61 0.61

Massgain:316 Repeller:27.78 MSsource:230
Massoffs:-11 Ionfcus:89.2 MSquad:150
Amugain:2223 Entlens:9.5 turbospeed:100
Amuoffs:122 Entoffs:19.33
Wid219:-0.048 EMVolts: 1576


h2o:0.35% n2:4.77% o2:1.43% co2:0.11%

Ramp criteria:
Ion Focus max: 90V using 502 ; EM gain 57358
Repeller max: 35V using ion 219; gain factor 0.47


Any advice would be very appreciated. Thank you.

[James_Ball]

I can't open the links through our work firewall so can't see the chromatogram, but what are the masses in the noise of the baseline you are seeing.

Remember, at 10:1 split that 100ng has become 10ng on column which is quite low especially for hydrocarbons on a 5973. The instrument can see quite low concentrations but only if there are no background ions that interfere.

You might want to try a splitless injection. Begin with a simple splitless hold for 1 minute then 20:1 after the hold to sweep out any high boiling contaminates. Reduce the splitless time until you see a drop in response then add back a little and you have it pretty much optimized.

[buggygcms]

I can't open the links through our work firewall so can't see the chromatogram, but what are the masses in the noise of the baseline you are seeing.

Remember, at 10:1 split that 100ng has become 10ng on column which is quite low especially for hydrocarbons on a 5973. The instrument can see quite low concentrations but only if there are no background ions that interfere.

You might want to try a splitless injection. Begin with a simple splitless hold for 1 minute then 20:1 after the hold to sweep out any high boiling contaminates. Reduce the splitless time until you see a drop in response then add back a little and you have it pretty much optimized.

The masses are 40, 44, 73, 207, 325, 340, and 42, with 40 or 44 as the base peak depending on the RT.
My injection was a splitless injection which was held for 1 min then purged 10ml/min (which was ~9:1). I will try a larger split ratio along with different hold times to see if that makes any difference. Thanks!

[cjm]

I assume you're doing a hot splitless injection with maybe a straight liner single taper with wool? What are your inlet temps? Abundance of 69 looks good, I like to keep under 600k. 502 looks high at 14%, I'm not sure I've ever seen our systems get that high. Not that you can't run that way, and not that it will help your chromatography.

What is your threshold set to? It may be too low

What is the sampling rate (2^n) what value of n?

Do you have any chromatography you can share and maybe a full tune report?

[dblux_]

... I need my instrument to be able to detect <100ng (with a clean baseline ideally) and I'm not sure what to do.
...
Any advice would be very appreciated. Thank you.

Change to SIM mode instead of SCAN.

[buggygcms]

I assume you're doing a hot splitless injection with maybe a straight liner single taper with wool? What are your inlet temps? Abundance of 69 looks good, I like to keep under 600k. 502 looks high at 14%, I'm not sure I've ever seen our systems get that high. Not that you can't run that way, and not that it will help your chromatography.

What is your threshold set to? It may be too low

What is the sampling rate (2^n) what value of n?

Do you have any chromatography you can share and maybe a full tune report?

My inlet temp is 250, and the threshold is set to 150.
My number in the sampling rate box is 2. I'm not sure it that is the value of n, or if it is the result of 2^n.

[James_Ball]

I can't open the links through our work firewall so can't see the chromatogram, but what are the masses in the noise of the baseline you are seeing.

Remember, at 10:1 split that 100ng has become 10ng on column which is quite low especially for hydrocarbons on a 5973. The instrument can see quite low concentrations but only if there are no background ions that interfere.

You might want to try a splitless injection. Begin with a simple splitless hold for 1 minute then 20:1 after the hold to sweep out any high boiling contaminates. Reduce the splitless time until you see a drop in response then add back a little and you have it pretty much optimized.

The masses are 40, 44, 73, 207, 325, 340, and 42, with 40 or 44 as the base peak depending on the RT.
My injection was a splitless injection which was held for 1 min then purged 10ml/min (which was ~9:1). I will try a larger split ratio along with different hold times to see if that makes any difference. Thanks!

40 is Argon and 44 is CO2, mostly coming from tiny air leaks or impurities in the helium itself. The rest probably are column bleed and should get larger as the temperature increases.

Does the 44 increase or decrease with increasing temperature?

If there is a quick decrease then it could be a leak at the MS interface, the vespel ferrules can shrink with repeated heating cycles of the oven and need to be tightened about 1/8 turn of the nut. I try to keep a small beaker with a few ferrules in the oven so they are pre-shrunk to avoid that problem when changing a column.

If you are already splitless I doubt much longer than 1 minute will make much difference, but you could also try a pulsed splitless injection which will transfer the analytes to the column faster and minimize the solvent expansion volume to give tighter peaks. Also if you are only looking at hydrocarbons with no worry about fragile analytes degrading in the inlet you can try raising the temperature of the injector 25C and see what happens. Also as above what liner are you using? If no glass wool, try adding some to help get the analytes into gas phase more quickly.