Chromatography Forum

This group upstairs has an Agilent 6890N with two inlets coupled to a 5975N MS detector. Currently, one of the inlets is directly interfaced with a catalytic pyrolizer so that they can analyze stuff inline (inlet connected to single column to mass detector). They would like to be able to use the second inlet to do manual injections of liquid samples and therefore connect both inlets to a single column, going to the single MS detector (I've seen a lot of information/posts about 2 columns or two detectors, but not much about this particular configuration).

Our initial idea was to use a metallic Y-connection (reusable and less fragile than the glass ones); we were told that our best bet was to use a Dean switch, but that we would also require an auxiliary EPC on the GC to control the flow through the switch, which means the cost is significant (well, for an academic lab with a very tight budget).

Just wanted to see if anyone had any thoughts on the best and cheapest way to accomplish this and/or experiences with such a setup.


Thanks!

I am currently doing this with two injectors, two columns into one mass spec. You should be able to do it using one column without all the extra switches and valves. I would use a 5m guard column from each inlet into the metal Y connector ( I use the MXT one from Restek) or a glass Y connector to mate the guard columns to the analytical column. The reason you want a good length of guard column is so you don't have problems with flow backing up into the opposite column.

Once connected, you set the flows so that the inlet you are using is providing a flow at 1 ml/minute and set the unused inlet to provide flow at 0.2-0.3ml/min. That way you are getting a good linear velocity through the column without overloading the detector, and the flow from the unused inlet keeps your analytes from flowing back up into that inlet.

If you want to get more advanced, you can switch flows rates after a few minutes so that the unused inlet is at 1.3ml/min and the other is at 0 which will allow a backflush flow to clean out the guard column of heavier compounds. You will have to experiment with the flow settings to see what it takes to get the forward flow at the same velocity and provide back flow in the opposite inlet if you really want to try this, the 1.3 and 0 were just examples.

You may also want to use guard columns that are slightly smaller than the analytical column so that the flow is more prone to follow the path to the MS and not back to the opposite inlet. When I did it with two columns I found I had to use a short section of 0.53mm guard at the MS interface for two 0.18mm analytical columns, this allowed the end of the two columns to be exposed to a vacuum instead of the pressure from the opposite column. When I first tried it with a 0.1mm transfer line, I had problems getting flows calibrated because there was a positive pressure at the Y union causing back pressure on the analytical columns.

Of course the most simple way to do this is to just put a nut with a blank ferrule in the unused column and switch the blank ferrule nut and the column nut between the two injection ports depending on which one you want to use at the moment. It requires you to keep track of which one it is installed in but is easier to set up. With the Y union you just switch methods to switch columns.

It is too easy to over-complicate this. Connect the column to the MS. At the inlet end you move the column to whichever inlet you want to use, and cap the unused one with a no-hole ferrule in an inlet nut. As long as the two inlets are the same you do not move the nut or ferrule on the column, just unscrew from one inlet and screw into the other. Job done.

Peter

Thank you both for the replies; I guess that how often they wish to switch inlet will be something to consider.

Peter, by simply switching it, wouldn't this allow air into the MS?


Thanks again

it does not take long to switch from one injector to the other.
and the air intake in the column is only caused by the vacuum of the MS.

this effect should be gone in a few minutes.

Peter, by simply switching it, wouldn't this allow air into the MS?

With 30m of capillary between the inlet and MS, it's not a big concern as long as you do it somewhat quickly.

Thank you all; really appreciate the input!

Thank you all; really appreciate the input!

Also to minimize problems with air just let the carrier flow for five minutes with the oven temperature set to 30C, this will flush out the air and not oxidize the stationary phase, and that little slug will not cause any problems at the detector.

Hi Trishia,

I would recommend a much cheaper option.
There are Special "two hole ferrules" available, where you can install two columns in one MS.

I am shure you will found this at different suppliers.

greetings

Hi Trishia,

I would recommend a much cheaper option.
There are Special "two hole ferrules" available, where you can install two columns in one MS.

I am shure you will found this at different suppliers.

greetings

But this needs two columns, and although I've never tried it I'm pretty sure that you can't fit two columns through the transfer line tip. And even if you could, unless you play around with pressures and flows it increases the total flow to the MS and makes it harder for the pumps to pull a vacuum.

Peter

Hi Trishia,

I would recommend a much cheaper option.
There are Special "two hole ferrules" available, where you can install two columns in one MS.

I am shure you will found this at different suppliers.

greetings

But this needs two columns, and although I've never tried it I'm pretty sure that you can't fit two columns through the transfer line tip. And even if you could, unless you play around with pressures and flows it increases the total flow to the MS and makes it harder for the pumps to pull a vacuum.

Peter

0.18 ID columns have an outer diameter near 0.4mm and a 0.53ID column has an outer diameter of 0.8mm so theoretically you could slide two 0.18ID columns through the transfer line (at least of an Agilent) at the same time. The problem would be that the two holes in the two hole ferrule would be wider than 0.8mm and I would worry that at the point where they meet the interface they would be crushed when the ferrule is tightened. Now if the interface was beveled it might work, but that would require altering the interface and possibly preventing the ferrule from sealing. The hole in the nut would also need to be enlarged to accept the two columns.

This setup would be far superior to the one I am running with the Tee ( I run an Rxi624SilMS and a Rxi5SilMS to do volatiles and semivolatiles on the same instrument). Flow is easy to control with inlet pressure switching.

It is an interesting idea, and I would jump in it if the interface is compatible. Maybe I will do some research on this in the future.