Methods to separate co-eluting peaks

[sprucebvoc]

Hej! I am new to working with GC-MS and have a question regarding separation of peaks.

Some background: I am working with BVOC emissions from spruce trees and want to analyse these. To do so I am using adsorbent tubes (collecting volatiles in field) and doing analysis with a TurboMatrix ATD650 coupled with a Shimadzu QP2010 Plus GC-MS system. I ordered and prepared some standards (Isoprene, monoterpenes and sesquiterpenes), which separetes fine and have nice peaks - except for the sesquiterpenes longifolene and beta-caryophyllene which is co-eluting in my chromatogram. So this means that I cannot identify these compounds in my samples, leaving me with three options:

1) Switch column (not optimal due to delivery times and cost)
2) Change the method for my current column (BPX5, 50m, 0.32 mm, 1.0 um)
3) Ignore and analyse the two compounds combined

So my main question is to get advice on what to do - and I also wonder if it actually is possible choose option 2 and change the method? I currently run the following program on the GC-MS: starting temperature at 40 C for 1 min, increase to 210 C with 5 C intervals and once there increasing to 250 C with 20 C interval and hold for 2 minutes. The total time is approx 39 minutes. The column flow is 0.71 mL/min.

Since I dont have that much support at our department (no lab manager etc) I hope some of you might be able to help. Thanks!

[benhutcherson]

Plugging your column dimensions into Restek's calculator gives an optimum flow rate of 1.7-2.6mL/min helium( https://www.restek.com/ezgc-mtfc )

I would actually try a flow rate closer to this value, assuming that your MS vacuum system can handle it.

Although it may seem counter-intuitive, in your case increasing the flow rate(or more specifically the linear velocity through the column) may well give narrower/sharper/better resolved peaks that do baseline separate.

A couple of other things-are you currently doing split or splitless injections? Even a mild split, if you are not doing one, can improve peak shape dramatically.

Second, I would seriously evaluate my column dimensions, or in particular the diameter. .32mm is awfully large for an MS column. In addition, smaller diameters tend to have higher theoretical resolutions for a given length. The increase is dramatic enough that you can/do typically use a shorter column and consequently can end up with an even shorter analysis time while still improving resolution. I don't routinely run anything larger than a .25mm, and run a fair bit in the .15-.20mm range. Something like a 30mmx.25mm is a good balance for a lot of analytes.

Finally, if possible, consider switching to hydrogen.

[rb6banjo]

I'm not sure, even if you do add 2D (heartcutting) capability to your system that you'll be able to separate these hydrocarbons. They're so similar. But, it has helped me many times in my career. You can probably add a Dean's Switch to your system. You probably want the identification using the MS for your analysis so to really make this work well, you'd probably need 2 MS detectors. But, if you can live with a flame detector on the DB-5 separation, you could make it work for not a lot of money. I just thought I'd throw it out there as an option. It may give you another idea.

In this example, camphor and menthol coelute on a DB-5 phase. A simple heartcut from the DB-5 phase to a wax phase (pretty much polar opposites) made the separation easy.

Precolumn sepn. with heartcut:



Analytical column (wax phase) sepn. of that cut:



The complexity of real samples has led me to never purchase a GC system that didn't give me the option for heartcutting. It's a very powerful tool. My system is an Agilent 7890/5977/FID (flame used to monitor the effluent from the precolumn (DB-5 sepn.). Shimadzu sells a Dean's switch. I've seen it applied in some of their app notes. Certainly you can cut the entire precolumn separation to the second column. You just get a mixed separation of a DB-5 followed by a Wax.

[James_Ball]

Another trick you can try is calculate the temperature at which the pair elute, then at about 20-30C lower than that add a 1-2minute hold in the temperature gradient then resume at the same rate. Sometimes that is enough to separate a close eluting pair.

Do those have the exact same spectra, or is there a mass in each that is unique? If you have a unique mass in each then you don't have to separate them completely to quantify them accurately.