Leaking Multi Dean's Switch

[Selterbude]

Hello everybody,

we are currently using a Shimadzu GCMS-Q2010 Ultra coupled with a Shimadzu GC-2010 Plus (TCD) connected using a Multi Dean's Switch also by Shimadzu. However, even if we set the valve to -110, which should close the valve and lead to no gas reaching the TCD, we still get a signal at the TCD. We are now trying different pressures for the first column's carrier gas, as well as different pressures at the APC controlling the Dean's Switch and different presures for the second column. My question would be, if someone has an idea as to which pressures seem appropriate or if you have any other input.
Thanks for reading

[Peter Apps]

Welcome to the forum

A Deans' switch is never closed, it always has gas coming out of both sides - the switching just moves the carrier gas loaded with analytes from one side to the other. When your switch is switched away from the TCD, the TCD should be getting only clean gas.

I have never seen a Shimadzu Deans' switch, and it could be that they are using some other kind of plumbing to achieve a similar effect

Presumably you are seeing peaks on the TCD when the switch is switched away from the TCD ?

What does the number -110 refer to and is it a pressure or a flow ?. Are you trying to switch between a second column and a detector ? - if so there will be a bias for flow to be towards the detector side where the flow resistance is lower. If you are switching between a TCD and an MS the vacuum in the MS will bias flow that way.

Peter

[Selterbude]

The -110 refers to a specific code you execute to toggle the Multi Dean's switch (there is supposed to be a valve in there somewhere that is triggered by this code). On top of that, we can adjust the flow. We are not switching currently, since even if we try to just send our gas flow to the MS, we see peaks at the TCD.

[Peter Apps]

Please post all the details of your column types and dimensions, flows and pressures. What are you trying to analyse ?

Do you see peaks on both the MS and the TCD or only on the TCD ?

Have you asked Shimadzu about this ?

Peter

[rb6banjo]

I have the Agilent version. I had to "balance" the system when I installed my columns. I first took the parameters that their "flow calculator" gave me but I obtained poor peak shapes, especially for the lighter materials, on the first column. My system is:

Injection is made on a 30 m x 0.32 mm x 0.25 µm ZB-5 (Phenomenex, going to switch and then the FID). I have the option to cut to a 30 m x 0.32 mm x 0.25 µm Simplicity Wax (Supelco) which goes into a mass spec. For this setup, the head pressure on the first column is 8.42 psig and the second column is 4.32 psig.

I set up a method that cut the entire volatiles profile to the second column. I then ran the same sample over and over, increasing the head pressure of the first column until I saw only a small burp of material in the flame for only the largest peaks in my chromatograms. Generally, I'm looking for trace components so as long as I get 100% of those materials cut to the second column, I'm happy.

Using this type of "balancing" procedure, I was able to obtain the best compromise between 100% cuts from one column to the other, minimizing the bleed to the monitor detector (in my case, the FID) and peak shape after injection. I analyze beer headspace using SPME and I only see a blip for ethanol in the FID trace when I cut the entire beer profile to the mass spec. (essentially a 60 m separation by the ZB-5 and then Wax is succession). Here is an example of my chromatography. Someone on the forum was having trouble with separating menthol from camphor (in a medicinal rub material) and I easily demonstrated how this system can solve the problem:

https://onedrive.live.com/?authkey=%21A ... ot&o=OneUp

I hope this helps. I have other heart-cutting setups on other instruments but those use an on/off valve that shuts off the flow to the monitor detector (again a FID) and then the "path of least resistance" forces the flow for the duration of the cut, to the second column. The EPC modules go slightly out of control momentarily but they recover quickly. I've never ever had problems with the accuracy of my cuts on those systems. They have an olfactory detector at the end of the second column and they are used for solving off flavor/aroma problems. Very powerful!! Good luck!

[Selterbude]

Thank you for your answers already! As samples we have used 1,5% CO2 in Helium and also pyrolysis of calcium oxalate. Our first column is a UltraAlloy column, while the second one is a Carboxen one. Right now we try to use samples that are seperateable by the Carboxen column (i.e. CO2, CO, etc.). We are experimenting with different pressures. Right now we are trying to use a head pressure of 108 kPa at the first column, different pressures at the Dean's Switch (we have tried anything from no pressure applied = APC off, up to 100 kPa applied). We have also tried different pressures at the second column (anything from 26 kPa to 216 kPa). The main issue is, that no matter if we close the valve at the Dean's Switch, the first column leaks into the second one and gives us a signal at the TCD. The signal does decrease though if we close the valve to the second column (but its still leaky). However, it is possible to not have any signal at the MS (after the first column) if either the valve is closed or the valve is open and the APC pressure is above 30-40 kPa. How do you usually cut to the second column? By turning on your APC at the Dean's switch or by closing a valve?

[rb6banjo]

As Peter mentioned, the flow to your detectors is never off in a switch like this. Think of it as a diverter valve. Depending on your settings (which direction you apply pressure at the switch), you push your analytes to one column or the other. This document describes the shimadzu version of the switch. You can see it's all about where you apply the pressure to push the analytes one way or the other.

http://www.ssi.shimadzu.com/products/li ... psMDGC.pdf

Note that on slide 6, a pressure is applied by the APC to ensure that 1) the second column has a nice flow of carrier gas but the path of least resistance is to Detector 1. This is the mode where you will see the majority of your analytes at Detector 1. When the APC pressure changes (slide 5, now through the fixed restrictor so the flow/pressure is less at that point in the switch), the path of least resistance is theoretically through the second column. In this mode, you should see the majority of your analytes at the second detector.

If your system is not "balanced" you might see bleed from one to the other - especially for your largest concentration analytes. You can see that for my system where I have 2, 30 m columns, the pressure difference is (8.42-4.32)*6.89 = 28.2 kPa. You will likely have to play with it a little to get it to work correctly. I will also tell you that I operate in constant pressure mode but in theory, it should work for constant flow as well. My GC's with the olfactory detectors operate in constant pressure mode so that's what I'm used to doing.

[rb6banjo]

What are the dimensions of your columns and the fixed restrictor you use to go from the Dean's switch to the TCD? My restrictor for the 2, 30 m column configuration is a 3.64 m piece of 0.25 mm deactivated fused silica. it is important to have a piece of transfer tubing that presents a flow resistance that is similar to the second column.

If your fixed restrictor is correct, I would try this. Set your APC to the pressure that gives you the flow you want through your second column. Set the system in cut mode (path of least resistance to the second detector). Then, set the inlet pressure on your first column to something that will ensure you are not in backflush mode (i.e., when you inject, your analytes will move out of the injector then onto and through the column). Now, keep injecting the same sample over and over but each time, increase the inlet pressure a little bit. Repeat this until you just barely start to see peaks at your first detector. Then, back it off just a little bit.

[Peter Apps]

Can I ask again; what are the dimensions (length and internal diameter) of the two columns ? The more you don't tell us the more we can't help.

Peter