Inlet Pressure Too Low

[Mendicant]

Hello there.

It's another great day here at Noob-Chromatographers-R-Us, and I'm trying to build a method mostly from scratch. I have a published method that I'm trying to replicate, but we're a little lacking in infrastructure.

We're trying to separate some big, organic compounds, but they tend to self-destruct at higher temperatures. The good folks we borrowed an SOP from suggest using a cold on-column injector or a PTV. Unfortunately, I have neither of those available for the ol' 6890 I'd like to run this on, so I am attempting to make do with a splitless liner.

BACK INLET
Mode: Splitless
Initial temp: 100 C
Pressure: 3.6 psi
Purge flow: 0.0 mL/min
Purge time: 0.00 min
Total flow: 4.2 mL/min
Gas saver: Off
Gas type: Helium

Unfortunately, when I try to run this, the pressure is not able to get above the 1.5 psi mark. At first I thought it might be a leak, so I replaced the septum, the O-ring, and checked the liner. Tried again and it would only get to the same pressure. Opened the oven, unscrewed the column nut, changed the ferrule, and cut/reinstalled the column, and tried again. Still the same pressure.

Tried going for a splitless mode, set the split ratio at 5:1, and it gets up to the requisite pressure just fine.

What I'm wondering is, is the liner I'm using too large to pressurize?



I believe the one in there may be an SGE model 092001 that someone elected to knock the wool out of at some point. I see from the description it is "4.0mm ID Spilt, with quartz wool." Does that mean it only works with split methods? If I try wedging something inert in there to obstruct the flow, can the pressure go up higher?

Many thanks in advance for all your suggestions.

[tkubowicz]

Hello

In your method, you set:

Purge flow: 0.0 mL/min
Purge time: 0.00 min

I don't know why it is like that but it's wrong. You need to set those parameters. Purge flow allows you to remove excess of your sample from inlet after specified time (purge time).

I'd recommend

Purge flow: 45-70 mL/min (try 60ml/min for start)
Purge time: 0.3 - 2 min (try 0.45 min for start)

Regards

Tomasz Kubowicz

[Bigbear]

What is your column dimensions? If it is a capillary column you will need more than 4 psi.

[Consumer Products Guy]

Isn't total flow level really low? Like except for the programmed time of the splitless part, the unit operates like in split mode.

[Mendicant]

Hello

In your method, you set:

Purge flow: 0.0 mL/min
Purge time: 0.00 min

I don't know why it is like that but it's wrong. You need to set those parameters. Purge flow allows you to remove excess of your sample from inlet after specified time (purge time).

I'd recommend

Purge flow: 45-70 mL/min (try 60ml/min for start)
Purge time: 0.3 - 2 min (try 0.45 min for start)

Regards

Tomasz Kubowicz

I agree that eventually I would want a purge flow. In my previous thread, I had actually discussed some huge solvent peak tailing from this phenomenon. viewtopic.php?f=2&t=27824

Unfortunately, these are some big organic compounds that I'm injecting, and I have no idea what temp the inlet needs to be in order to volatilize them. Nor do I want to overestimate and incinerate them all before they make it to the detector. The SOP I'm reading says to use either a cold on-column injector or a PTV, but alas, I have not the items necessary to set up either of those systems on this machine at this time.

My plan was to use a splitless liner and inject normally, and then slowly raise the inlet temperature to see when they start and stop making it to the column. But if I set up a purge flow now, I'll lose most of the material and won't have the possibility to see the compounds on the detector side.

[Mendicant]

What is your column dimensions? If it is a capillary column you will need more than 4 psi.

ID = 0.320 mm
Length = 5.5m

The column I initially cut this length from was an Agilent 123-5012, which is manufactured as a 15m column. I tried adjusting the length in the chemstation software, but it would only let me cut 5 meters off, so it thinks it has a 10m column installed. I still need to figure out how to calibrate it such that it knows it is actually only a 5.5m column hooked up.

[Mendicant]

Isn't total flow level really low? Like except for the programmed time of the splitless part, the unit operates like in split mode.

I don't know; is it really low?

What would be really low and what would be really high?

What do you think I should try changing the flows to?

[Bigbear]

Are you using Chemstation D or later? there is a function where you set column length by timing an retained compound.
It's been years since I've used a 5890, but can it be configured for H??

Why would you want to use such a short column?
The point of GC is to have the inlet hot enough to flash boil everything, while having the oven cool enough to condense the compounds into the liquid phase.

[Mendicant]

Are you using Chemstation D or later? there is a function where you set column length by timing an retained compound.
It's been years since I've used a 5890, but can it be configured for H??

Why would you want to use such a short column?
The point of GC is to have the inlet hot enough to flash boil everything, while having the oven cool enough to condense the compounds into the liquid phase.

It is a 6890, not a 5890.

Chemstation version C.01.06

I see there is some feature that can set the column length by retention time, but when I was attempting it it would not accept values that calculated the column being shortened by more than 5m.

The method I am referencing specifies a length of the column to be 4 to 7 meters in length. Toward the 7 meter side of things, these compounds start to degrade. "Thermal degradation may affect analysis by diminishing recovery on C58 an [sic] C60, if this happens a
shorter column should be used ) – Normally a 5 m column leads to acceptable results." For purposes of discussion, the C58 and C60 being quoted here refer to derivatized diacylglyceride and triglycerides.

I agree that the point of the inlet needs to be hot enough to flash boil both solvent and analyte. What I do not currently know is what temperature will accomplish this without incinerating these organic compounds.

My goal is to find out if it is possible to achieve this using a splitless method. And I am unable to start this process of deductive reasoning until I actually inject something. Which the autosampler is not currently letting me do, because the pressure in the inlet is too low, causing a shutdown. I need to figure out why the pressure is so low. I have addressed some potential leak sources, and am currently questioning the suitability of the liner, or if my flows are set up correctly.

Again, thank you to everyone so far for the input.

[Mendicant]

UPDATE:

I have additional data that may be relevant.

I do not believe the liner to be the cause of the problem.

I have tried liner SGE 092007, which is a split liner, and Agilent 5062-3587, which is a splitless liner. With both liners, the maximum pressure the inlet can attain in splitless mode is 1.5 psi.

However, with either type of liner, when I switch to split mode, it has no problem getting up to whatever pressure is needed. This in itself is not helpful to me as I need all of the analyte to make it over to the column, but it may be helpful for troubleshooting.

Why would an inlet be able to reach a higher pressure in split mode and not be able to meet the minimum pressure in splitless mode?

[Mendicant]

SOLVED!

tkubowicz hit it on the nose.

I called up Agilent technical support and here was what I got, aside from terrible hold music-

The purge flow needs to be set. Because there is no split flow and the column flow is quite low, with no purge flow, the total flow is too low. Much as Consumer Products Guy was saying.

Because the total flow is so low, the GC cannot accurately control it. This fact seems counterintuitive to me. I would think that it would be when the flow is extremely high it would be hard to control, and it should be much easier to control when the flow is lower. But this is why I am here tinkering around with a method, and they get paid the big bucks.

Now I need to bid you adieu and try to get the FID working...