One GC Two MS?

[HeadspaceLady]

Hi, i am looking for information to see if it is possible to get an agilent gc 6890 work with two MSD at the same time.

My thought was one gc, 2 columns and 2 MS. one MS on the left side and on the right side of the GC.

More questions:
(1) I saw pictures that 6890 can connect an olfactometer out from the left side of the oven and the output line which looks like a headspace transfer line. (I found a picture from this website http://www.mdgc.com/company.html )
I would like to know if i need to remove some parts of the gc in order to making a hole.
(2) if i use the transfer line to connect with the MS, the printed circuit boards can stand ~200℃ without problem ? I am not sure the max temp for the line as well.
(3) how about the transfer line out from the top of the oven?
(4) Does 6890 have enough cable ports for #2 MS?
(5) Can a regular version of chemstation control 2 MS at the same time?
(6) Any problem comes accross on your gc parts (eg over heat, software?) ,when you use gc-o?
Many thanks!

[James_Ball]

You would be using a 5973 or 5975 mass spec I would imagine, since those work normally on the left side(facing the oven door). The AUX heater there controls the interface temperature and defaults to Thermal AUX#2, there is also a Thermal AUX#1 which I am guessing could be configured for the interface heater for the second mass spec, but you still need some type of controller and power supply for a transfer line. If you used a transfer line from a purge and trap, that would probably work, but the temperature for any transfer line that long will be limited and you might have to worry about cold trapping if it won't go up to about 250C.

You will need a remote cable to connect the three instruments together for start stop signals, but that could be fabricated similar to the one used on a purge and trap or headspace unit, I just don't know if the second mass spec is going to accept the signal from the GC and first mass spec. You would probably need to set it up as a second instrument using Other/None as the Sample Inlet and External Source as the Injection Source. You would then have to run two sequences at the same time with the second instrument taking start commands from the GC on the first instrument. You would need to use two instruments because there is not place in the MS Configuration for anything but a GC and MS, the second instrument would be configured as MS only. To make it work you would need to start the second instrument then start the first since the second would be a slave to the first, otherwise the first could inject before the second is ready. At worst case, you would have to control the second with a second PC, but I think you can run two instances of MSDChemstation at once if your PC can handle it.

[HeadspaceLady]

James, thank you for your information. It gives me hope to keep this project going!! You are right. They are 5973.

The gc system has a g1888 headspace sampler, so do you mean i need to plug a remote start cable into pc, gc, headspace and 2MS to make it work? How do you call this cable (APG remote start cable 1 to 5?), or i am still able to use the y cable?

How about the flow control? Would that be a problem if i dont have a EPC splitter?
I plan to use db624 30x 0.32mmx 1.8um and innowax 20mx0.18mm x0.18um.

Thanks.

[rb6banjo]

This is a new one on me and it's interesting. What is your application? Can you provide some more details?

[HeadspaceLady]

This is a new one on me and it's interesting. What is your application? Can you provide some more details?

Our place has a not working MS, so i hope i can make use of it to try on some residual solvent methods ( eg USP 467). It could save time and headache especially on out of spec. sample in one injection.

Wish me luck to get approval from my boss to work on the challenge.

[James_Ball]

Are you wanting to do a confirmation or actually do two samples at the same time?

An easier way to do the confirmations but would require two injections would be to interface both columns to one MS, and inject on the first then the second column, having normal flow on the active column and low flow on the inactive column. Requires two methods but can work.

As for the APG remote, yes you need it connected to all the units. If not connected to the second MS, it will not receive the start signal to make it acquire data at the same time as the first.

[rb6banjo]

I was thinking like James. His is probably the more practical approach. If it's a "not working MS", how can you use it?

[HeadspaceLady]

James, i like your technique 2 columns to 1 MS ! I will see where i can use it. I asked about the not working MS already. It should be fine after putting back all the missing parts. Its arms and legs are inside different MS as parts now.



The GC is 6890A. It has two APG start remote connectors. I want one to connect to MS#2, one connects to MS#1 and headspace.  Can two MS  start and chemstation collects data when the headspace injects a sample in this setup?

Pic is from this website http://gcheadspace.blogspot.ca

What do you think if i would like to see one injection only, and then sample go through 2 different polarity columns to the 2 MS at the same time. I thought when sample is out of spec, i can confirm the problem compound right away. And then i just prep the sample duplicate to conform the concentration. I assume it could solve the co-eluting peaks, save sample preparation time, supplies and shorten the run time. It also can have less false positive. That means we can have more confidence result.

I am still in collecting information stage. I do this on my own time for work, so it will be a long way to go. I am really happy i can find support here for my crazy idea.

Hope to hear the good news about the cable connection. Thanks!!

[James_Ball]

Im not sure if you have to have them all on one cable or can use the two connections on the GC, but as long as all the parts can receive the same Ready and Start signals it would work.

If the two columns have completely different elution orders and with the same oven program you would not have any co-elution of the same compounds, there might be an even easier solution to your problem I just thought of. One HS, one GC, two columns, one MS.

Headspace connected to inlet, then splitter to the two columns, then another Y connection at the ends of the columns into a transfer line into the MS. If you can run the columns at about 0.75ml/min each the MS should handle the total flow. You would simply have two peaks for each compound at different retention times on the same chromatogram. It would be similar to doing 2D-GC but without the Dean's Switch. I don't know that it has been tried yet, so maybe it would be something you could write up.

[HeadspaceLady]

Thumbs up! James!! Thank you very much!

The new method added some more compounds, so i am still not sure if it works on another column. If the overlap peaks on the #1 column can have better peaks on #2 column and there are not many interference peaks in most of the samples, it would be perfect. But unfortunately, the coming months are going to be very busy for our company. My boss said i should focus on my routine work. He also said triple quad is better than 2 columns methods, but we have no plan to buy another one anytime soon. So, i don’t know when i am able to try anything or buy y-splitter. Maybe someone could share their results after trying out any of the methods. ^_^ keep going!

[James_Ball]

Thumbs up! James!! Thank you very much!

The new method added some more compounds, so i am still not sure if it works on another column. If the overlap peaks on the #1 column can have better peaks on #2 column and there are not many interference peaks in most of the samples, it would be perfect. But unfortunately, the coming months are going to be very busy for our company. My boss said i should focus on my routine work. He also said triple quad is better than 2 columns methods, but we have no plan to buy another one anytime soon. So, i don’t know when i am able to try anything or buy y-splitter. Maybe someone could share their results after trying out any of the methods. ^_^ keep going!

I currently run the two columns into one MS(7000 QQQ) but one column is from the injection port and the other column is from a purge and trap so I can do either volatile or semivolatile work without changing columns.