Mass flow / Back pressure

[pueblito]

Dear friends.
I would like to know what´s the difference between a mass flow controller and a backpressure regulator, and why does HP5890 use both?.
Second, my 5890 packed column head pressure regulator is not working and I´m planing to perform VOCs using this injection port, for the moment I only have a Kofloc model 2203 mass flow controller, Could I use this one to desorb VOCs from a Purge and trap?.

Thank you very much.

[gpronger]

This is a bit difficult to explain, but the 5890 inlet is basically a controlled leak, with what flows through your capillary the "leak". The mass flow controller, maintains a constant flow of carrier into the inlet. the back pressure regulator maintains a fixed pressure through-out the inlet system. This back pressure controls the flow into the column (ignoring the effects of the temperature in the GC oven causing changes in resistance to flow in the column).

Lets pretend that there is 104 ml/min flowing from your mass flow controller into the inlet.

There are 3 ways for it to leave, one the column, two the split vent, and three the septum purge vent. The line from the split vent goes to the solenoid valve. The line from the septum purge goes to a tee with one arm connected to the septum purge controller, the other side of the tee goes to the solenoid valve.

The solenoid valve allows the flow in the inlet to exit either via the split vent line or via the septum purge line. Between the exit of the solenoid and the split vent is the back pressure regulator.

When you're in "splitless", the flow is leaving the mass flow controller and entering the inlet. The split vent is turned off, so that the flow goes across the top of the inlet leaving via the septum purge line through both the septum purge vent as well as the septum purge vent. Because your injection with a syringe is "below" the flow racing over the top part of the inlet it "sees" mostly the flow going into the column so that your sample goes (mostly) in that direction.

When you go from splitless to split, the solenoid valve switches so that the flow going into the inlet now sweeps down through the inlet removing any remaining solvent and analyte still hanging around.

So, lets say (again 104 mLs/min in), a septum purge flow of 2 mLs/min and a split flow of 100 mLs would have 2 mLs/min going into the column. This would be a 50:1 split, if you made a split injection. In the splitless mode, the 104 mls/min goes in, 102 mLs/min goes across the top and down the septum purge line, and 2 mLs/min is heading down the middle of your liner and onto your capillary column. when it switches, then all that is going down the septum purge line is the two mLs/min exiting the septum purge, the 102 mLs/min are going down the middle of the inlet, 2 mLs/min into the column and the remaining 100 mLs/min out the split vent.

hope that helps.

Greg

[pueblito]

dear gpronger.
many thanks for your reply, I finally understand this sistem "mass flow-backpressure" regulated.
But I have another question. What do you think about desorbing VOCs from a purge and trap using only a mass flow controller?

thank you.
bye.

[gpronger]

Interesting question, because I stopped my previous response in talking about since it was a not quite what you were asking about. In fact though, one of the problems with how the inlet is configured is it impacts using a purge & trap on the instrument and probably also a headspace unit (I've not set one up, but I would suspect that they're interfaced to the system the same way as a purge & trap).

If you think about how the inlet works, at all times, we have (using my previous example) 104 mLs/min going into the inlet. We switch from split to splitless by switching the flow from going down through the inlet (split mode) to across the top of the inlet (splitless mode). Now, with a syringe injection, the sample does not "see" the flow across the top so we get most of the injection deposited into our column. The problem arises when we are introducing the sample as part of the flow into the inlet. When we hook up a purge and trap to the GC, we typically take the flow from the GC to the back of the purge and trap, and then hook the purge and trap up to the inlet where the carrier gas normally enters the inlet. So, with how our inlet works, we're effectively always in the split mode regardless of what the valve is doing. The analytes that are coming in are simply going to follow the flow, even when we're in "splitless" they'll simply flow out the septum purge line along with everything else.

So, what to do?

1) You can hook the transfer line up directly to the column (there are panels on the top of the 5890 GC you can unscrew that allows you to pass the line through). You can then use the mass-flow controller to maintain the flow. The problem that is common with this configuration is the peak shapes of the early volatiles, particularly the gases.

2) You use Greg's trick. The predecessor of the 5890 was a GC called a 5880 (I'm dating myself here, though I also used the 5880's predecessors), which had a very similar inlet, but when you were in the splitless mode, the flow to the inlet was only what was going into the column (and also out the septum purge). The problem HP (now Agilent) had was flashback. The liners they used with the instrument were very narrow, and with the very low flow, you had a lot of material flashing up into the top of the inlet and actually the carrier gas tubing. Now, for volatiles we're not making a liquid injection so that's not a problem. the only significant difference between the inlet on the 5880 and the 5890 was how the inlet was plumbed.

This diagram is how your 5890 is most likely plumbed:



You need to convert it to:



(I was in the midst of trying to use words to explain this and came to the conclusion that it was going to be much easier to draw a diagram.)

What this configuration does, is that when the instrument is in splitless, the only flow entering the inlet is going on column or out the septum purge line. The flow that ends up going out the split vent is diverted around the inlet.

Assuming you're using a megabore type column, you should have sufficient flow for desorbing the trap. Typically I would like at least 10mLs/min to desorb the trap.

The benefits, are that you should get much better peak shape for the gases, and when the inlet switches to split, the increased flow will work to sweep moisture and heavier organics out your split vent.

In doing this, if you're unfamiliar with how the lines are hooked up in a 5890, Agilent (or HP) uses connectors that are plastic (usually white) nuts that need to be only finger-tightened. However, they use a small rubber o-ring which has the annoying habit of being stuck in the hole when you unscrew the line, or rolling off to hide place under the instrument so that when you put it back together you have a massive leak. My point is simply, double-check that you have only one o-ring (that one isn't stuck in the hole) and or if you think it's stayed in the hole, it didn't just roll away and hide under your instrument.

Take a look at the diagrams and feel free to email me back to better understand what I'd trying to explain.

[pueblito]

Thanks a lot gpronger.
good bye.