8260 missing early eluters 5890

[Balderquell]

Hi!

Short story: all gases and compounds even up to benzene are non-detect at any concentration level. Peak shape gradually improves from heavy tailing to very sharp throughout the run.

Longer story: this lab has the instrument set up a little differently than what I am accustomed. Using double tapered splitless liner and running totally splitless, their chromatography has been ok for gases until recently.

I removed the liner and it was packed to the gills with septa shavings and replaced with another double tapered splitless. I was hoping this would solve the problem but it is worse now by a lot.

Direct injection produces good chromatography if I turn split on after 1min. Normal sample purge/desorb produces the problem. I've exchanged purging units with known quality equipment. Same result.

I've always used a narrow bore split liner and a split ratio for volatile analysis. So I'm treading into new territory here =)

Any thoughts on cancellation of error with the septa shavings in the splitless liner?

[Peter Apps]

Unless you give us full details of the operating conditions any response cannot be anything other than guesswork.

Peter

[AICMM]

Balderquell,

1) Try split under the same conditions and see what you get. I have a long explanation but it is better to actually collect the data and see what happens.
2) Septum shavings probably provided vaporization surface area.
3) What is the total flow across the trap in desorb?
4) Have you trimmed the front of the column lately?

Best regards.

[Balderquell]

5890 GC 5972 MS
60M x .25mm ID column in known working condition. Currently low bleed.
Liner: Split Liner 1mm x 6.3 x 78.5 IP deactivated from Restek Cat# 20973 (installed today with much improvement in chromatography)

Run conditions at the moment:

Start temp 35C
Hold 4min
Ramp 9C from 35C to 100C
Ramp 15C from 100C to 250C
Flow 1.10ml/min
Split 50:1
Solvent delay 4.4min: first peak should be about 5.5min

OI 4560 concentrator w/ trap K

Purge time: 10min
Purge temp: 30C
Desorb Preheat 245C
Desorb 250C
Bake 260C
Pressure during Standby: 20psi
Pressure during Purge 12psi (seems too high)
Pressure during Desorb 12psi (also seems high)

Installed a narrow bore split liner today and overall things are looking better. However, sensitivity in general is still much diminished compared to history. I'm concerned about the purge unit pressures. I can turn the purge pressure dial on the 4560 completely off and see no bubbles during purge; but, the pressure reading remains constant. I can also turn the pressure dial up and see no change in the pressure reading. I've tried two different 4560s with the same result. Both do not have this pressure problem when connected to another GC.

I'm worried about an obstruction related to the injector port plumbing. Direct injection produces somewhat improved chromatography over purge but not by a substantial margin. Still some gases are non-detect at 25ppb while heavier (benzene +) compounds are sharp and detectable below 5ppb. I suspect septa shavings may be partially blocking split-vent, septum purge, or the transfer line connection. But I want to rule out method optimization before I start disassembling.

Internal standard responses are identical to historical performance when purged.

Standards have been prepared from fresh stock and are currently within expiration.

[Peter Apps]

What is the inlet temperature on the GC ?

Septa shavings come from syringe injections, and yet you have a purge and trap sampler which suggests the alarming prospect that the inlet liner was in the GC long before the purge and trap was coupled up. If even inlet liners were not getting changed there could be all sorts of other neglected maintenance, cleaning the MS source springs to mind but the problems that you report with the pressure suggest that something is also wrong with the plumbing in the P&T and or the inlet, as you suspect.

The heavier compounds producing sharper, more detectable peaks is very likely due to thermal focussing on the column salvaging inlet bands that are much too wide. If you run with high concentrations of the light compounds do you see wide (and tailing) peaks for those compounds ?

How is your P&T connected to the GC inlet ?- plumbed into the carrier gas feedline, or through the septum ? If the latter there might be a bit of septum stuck in the end of the transfer line and slowing down transfer to the inlet.

Step 1 is to measure some flows (with a flow meter not the front panel readouts on the GC) at the split outlet, septum purge, transfer line etc. I suspect that some dismantling and unblocking is going to be necessary.

Peter

[Balderquell]

It's always been used for purge and trap. The direct injections are for daily tuning purposes. At one or two injections a day, previous liner could've been in there for years from the look of it. Baseline is up and there is some evidence of silicon based contamination in the runs.

Mass spec has been recently serviced and source has been recently cleaned. Tuning regularly passes.

Inlet is 200C.

P&T transfer line is plumbed into the injector weldment. I've cleaned and inspected the transfer line for breaks/kinks and looks fine. I'm concerned there might be some septa particles or gunky build up in the weldment. It's difficult to access the transfer line opening or the septum purge opening in the weldment, however, to see if there is any blockage. But I can't confirm it's condition or clean it. I'm debating replacing with a new one.

While I had it opened, I cleaned and replaced some nuts and o-rings on split solenoid. It was filthy but didn't produce much change in chromatography.

[Peter Apps]

It sounds as if the split line and/or filter might be clogged - have you checked the split flow ?

AICMM asked about column maintenance, with the neglect that you have inherited there is a chance that there is enough gunk on the column to cause band spreading. What is the column history ?

Peter

[danicrd]

try to replace OI trap with new one and see if response improve. 12 psi of purge pressure is too high

[Balderquell]

Trap was new. The pressure gauge appeared to be faulty and was giving a false reading. I could disconnect the pressure gauge from the OI entirely and it would still read about 15 psi.

I was able to resolve this problem for a few weeks. And most of the maintenance was performed on the GC--new column, cleaned/deactivated injection port, and cleaned solenoid and associated plumbing.

[danicrd]

Sometimes also a new trap can be faulty... if you can try a certain good trap (also from another istrument)
Another common error with OI is during trap installation..sometimes the trap was installed at wrong side.... check installation direction

Good luck

[Balderquell]

An update and a new problem with this instrument.

The previous problem with this instrument has been mostly resolved. Peak shapes for gases and other compounds are narrow and symmetrical and with a good response. Replaced with a new Tekmar, new weldment, new column.

It ran for about a month until the RF for 1,1,2,2-tetrachloroethane dropped and has remained persistently below 0.300. Bromoform is running about 0.165 but is w/in limits. The method suggests activity as the culprit.

All other CCC and SPCC criteria are easily satisfied.

Method information is still roughly the same as before. The new Tekmar is programmed similarly (replacing the old OI unit dramatically improved chromatography).

In the time that the instrument was briefly working, few samples actually ran. So there hasn't been much opportunity for sample filth to work its way into the system.

On the assumption that the problem is related to activity, injector and column maintenance has been performed. Split line and solenoid have been cleaned. Heated transfer line is new and unused. Column is 3-4 months old and has been clipped. Source has been recently cleaned.

Stock standards are fresh from the manufacturers and freshly prepared. Both sources produce the same response factor issue.

The baseline is relatively spikey so there could still be some build-up somewhere. I've cleaned or replaced about everything I can find.

Any thoughts?

[AICMM]

Balderquell,

If you are still running splitless, most of your sample is being thrown out the top. Typically in this scenario, very little of your sample goes down the inlet so you get really poor response. What I would suggest is decreasing your split flow down to something more like 25:1 and running split all the time (Purge A on.)

Best regards.

[mckrause]

I see I have to start repeating my mantra here - you must use split mode for gas phase injections. AICMM is correct - about a 25:1 split. We tune our split ratio using vinyl chloride response.

Your ramp rates are really, really low - might want to ramp the GC significantly faster, plus decrease your initial hold time. You'll get sharper peaks (better sensitivity).

What is your desorb time? Excessively long desorb times (for reasons unknown to me) lead to loss of sensitivity and analyte.

Your desorb flow needs to be at least 10 ml/min to efficiently get the analytes off the trap.

[Balderquell]

Desorb time is 1min.

I've tried splitting at 20, 25, 50, etc. No improvement for 1,1,2,2-TCE. Chromatography is otherwise good.

Increased ramp rates and it helped peak shape and sensitivity of the gases quite a bit. No improvement for 1,1,2,2-TCE, however. Response factor is still at an average of 0.25.

I've tried various new traps.

[AICMM]

Balderquell,

Sorry to be a pain about this, but have you actually switched to split at the different flow rates you mention? It is not clear from your last post. If not, then you will still not be getting the optimal mass transfer onto the column.

Best regards.