Toluene repeatability on HS-GC-FID

[cwp]

I am having trouble calibrating Toluene for a residual solvents method. The peak area for toluene is decreasing from injection to injection on triplicate preparations of the same concentration (peak area goes from 11.4 to 3.1 to 1.6 for each injection of a separate vial of the same concentration). However, my internal standard (Acetonitrile) and Hexane peaks are consistent from injection to injection. I thought maybe Toluene was degrading, but I observed the same drop in area if I run the sample immediately after preparing it. I will detail exactly what I'm doing and GC conditions below:

I prepared a stock solution of 1% Toluene in Hexane, then diluted down to 0.001% into Hexane. 2 uL of 1% aqueous Acetonitrile was added to a 20mL headspace vial as an internal standard. Then 2 uL of the 0.001% solution was added and the vial was immediately sealed (manual crimp, cap does not turn when twisted). This was repeated twice more, to have a total of 3 vials of the same calibration level. Peak areas for Acetonitrile and Hexanes are <5% RSD over the three vials, while the peak area for Toluene dropped from 11.4 to 3.1 to 1.6.

Headspace Conditions:

• Instrument: Tekmar 7000
  Platen: 140 C
  Platen Equil: 0 min
  Sample Equil: 20 min
  Sample Loop: 160 C
  Transfer Line: 160 C
  Mix: off
  Vial Pressure: 20 PSI
  Pressurize time: 1 min
  Pressure Equil: 0.25 min
  Loop fill time: 1 min
  Loop Equil: 0.25 min
  Inject time: 0.5 min

GC Conditions:

• Instrument: Agilent 6890
  Inlet Temp: 250 C
  Inlet Liner: 1.0mm ID straight liner
  Split Ratio: 10
  Linear Velocity: 41 cm/s, constant flow
  Column: Rxi-624 Sil MS 30m, 0.25 mm ID, 1.4 um
  Oven Ramp: 35 C hold 1.5 min, 25 C/min to 300 C, hold 300 C 1.5 min
  Detector: FID
  Make up flow: 45 mL/min
  Hydrogen flow: 40 mL/min
  Air flow: 450 mL/min
  Data Rate: 20 Hz

[Peter Apps]

Is this lack of repeatability repeatable - how many sets of three have you run ?

Peter

[cwp]

Is this lack of repeatability repeatable - how many sets of three have you run ?

I've tried it four times - 2nd try I used acetone as diluent with same results, 3rd try I used 0.01% toluene solution instead of 0.001% and same result, fourth was with 0.001% toluene in water and same result.

[cwp]

I should also say, I was able to successfully calibrate 6 other compounds without encountering this error

[Peter Apps]

Just to clarify what you are doing.

If you prepare three vials as triplicates and run them one after the other in a sequence the first has a much higher area for toluene than the next two.

If you prepare one vial, run it, then prepare another one, run it, then prepare another one and run it, the first has a much higher toluene area than the second and third.

What kind of septa are you using ?

Try running a blank (i.e. everything except the toluene) then three standards, what is the pattern of peak sizes from the four vials ?

How is the headspace transfer line connected tot he GC ?

Check the inlet liner for dirt.

Peter

[cwp]

Just to clarify what you are doing.

If you prepare three vials as triplicates and run them one after the other in a sequence the first has a much higher area for toluene than the next two.

If you prepare one vial, run it, then prepare another one, run it, then prepare another one and run it, the first has a much higher toluene area than the second and third.

What kind of septa are you using ?

Try running a blank (i.e. everything except the toluene) then three standards, what is the pattern of peak sizes from the four vials ?

How is the headspace transfer line connected tot he GC ?

Check the inlet liner for dirt.

Peter

Yes, your clarification is correct.

The septa on the vial is silicone/PTFE (Restek Cat# 21763).

I will try running a blank (with hexane), then three toluene standards diluted in hexane and update with results in a bit.

The headspace transfer line is connected directly to the carrier line in the inlet manifold (carrier line was cut and a low volume dead union was installed a couple centimeters from the inlet). The union has been wrapped in insulation to minimize the cold spot.

I will check and clean the inlet before running the blank and standards again.

[Bigbear]

I suspect your temps. I'm under the impression you should'nt boil analytes. I would try:
seting platten to 60-70
lower your 7000 oven to 100 and the GC injection port to 120.

Also your loop fill time may be too long. Shoot for 3-5 loop volumes calculate time using loop volume and fill flow rate.

[Peter Apps]

I suspect your temps. I'm under the impression you should'nt boil analytes. I would try:
seting platten to 60-70
lower your 7000 oven to 100 and the GC injection port to 120.

Also your loop fill time may be too long. Shoot for 3-5 loop volumes calculate time using loop volume and fill flow rate.

Hi Bigbear, if I understand correctly the sample volume gives complete evaporation. Maybe cwp can confirm.

Peter

[cwp]

After trying what Peter suggested (cleaning inlet and running blank then standards), I am still observing the same issue with toluene, but not with any other analytes.

I ordered another reference standard from a different manufacturer just to see if it's an issue with the source. It won't be here for a day or two but I will update again after trying the new standard.


Hi Bigbear,
The method is a full evaporation technique, instead of a headspace equilibrium, so the idea is to completely volatilize the sample.

[Peter Apps]

After trying what Peter suggested (cleaning inlet and running blank then standards), I am still observing the same issue with toluene, but not with any other analytes.

I ordered another reference standard from a different manufacturer just to see if it's an issue with the source. It won't be here for a day or two but I will update again after trying the new standard.


Hi Bigbear,
The method is a full evaporation technique, instead of a headspace equilibrium, so the idea is to completely volatilize the sample.

Did you see any peaks in the blank ?

You may have uneven heating of the sample platen. Did you run the blank plus 3 standards as a sequence ? Which holes in the platen are the vials dropping into ? Is this the same for your two test schemes ? Which holes get used can depend on the sample equilibrium time in relation to the GC run time and the headspacer cycle time.

Peter

[cwp]

Did you see any peaks in the blank ?

You may have uneven heating of the sample platen. Did you run the blank plus 3 standards as a sequence ? Which holes in the platen are the vials dropping into ? Is this the same for your two test schemes ? Which holes get used can depend on the sample equilibrium time in relation to the GC run time and the headspacer cycle time.

Peter

I see hexane peaks and a small isopropanol peak (contamination from cleaning the benchtop before prepping the blank?) but nothing else.

I ran the blank plus standards as a sequence, but the GC cycle time is roughly equivalent to the platen equil time, so only 1 platen hole is being used at a time. (Vial is injected, returned to carousel, and next vial lowered into platen).

[Peter Apps]

As a trouble-shooting excercise; extend the time between the vial leaving the platen and the next one going in. Does this increase the peak area of the second vial ?

You did four series of three - what was the time lapse between the series ?

Peter

[RogerBardsley]

Hi cwp,

Have you tried adding the internal standard, acetonitrile, to the hexane stock? This would check the hexane sample introduction into the headspace vial versus the internal standar in water. Another suggestions is to prepare the internal standard in hexane also.

Best Regards,
Roger Bardsley
Applications Chemist
Teledyne Tekmar

[mckrause]

This is certainly not classic headspace. You have gained no advantage whatsoever over directly injecting the toluene in hexane standard, you have introduced several unnecessary variables and you spend too much on vials and caps.

If you insist on doing a direct injection method using headspace then you should use partitioning to your advantage. Add 2 mL of saturated salt solution to your vial prior to introducing the acetonitrile and hexane standards. It will buy you a couple of things. One, you'll be closer to true headspace analysis. Two, it will keep the toluene from dissolving in the PTFE liner of your vial septa.

Drop your temp to 70C; you want to run headspace well below the boiling point of the bulk matrix.

Good luck!

[Peter Apps]

This is certainly not classic headspace. You have gained no advantage whatsoever over directly injecting the toluene in hexane standard, you have introduced several unnecessary variables and you spend too much on vials and caps.

If you insist on doing a direct injection method using headspace then you should use partitioning to your advantage. Add 2 mL of saturated salt solution to your vial prior to introducing the acetonitrile and hexane standards. It will buy you a couple of things. One, you'll be closer to true headspace analysis. Two, it will keep the toluene from dissolving in the PTFE liner of your vial septa.

Drop your temp to 70C; you want to run headspace well below the boiling point of the bulk matrix.

Good luck!

While total evaporation is not strictly "headspace" it is a well established way of analysing volatiles from matrices that cannot conveniently be extracted, or injected as solutions.

In this instance the volatile standards are a stepping stone towards residual solvents in a matrix which cannot be injected as such without crudding up the inlet. Doing a total evaporation "headspace" analysis leaves the crud behind in the vial.

Peter