I've been trying to improve overall sensitivity using large volume injection (LVI) techniques on a Saturn 2000. Is there a book or article out there that goes into details wrt developing LVI's?
Some specific questions:
Is the goal to evaporate the solvent while the extract remains on the needle?
If not, why not fast inject (~5-10ul/sec)onto a wad of fused silica in the injector, remove the needle, and let the solvent evaporate from the fs?
Should column temp remain low thru the heating cyle of the injector, or can the column program start at the same time as the injector?
What are guidelines to use for rate of injection? Solvent dependent?
I'm getting 6 to 12 sec RT variations in early eluting peaks, using LVI methods I've developed. Suggestions for reducing RT variations?
Seems like someone could write a book on LVI. Anyone know of such a book?
Thanks for any help.
By Anonymous on Monday, December 11, 2000 - 01:06 pm:
I played around with LVI for awhile. I wanted to perform a large injection of wastewater directly. Unfortunately, water is such a poor GC solvent that I could never get it to work reproducibly. The only references I could find were application notes that came from the vendor. My LVI is a Gerstel (I also have the Agilent/Gerstel counterpart). Both are gathering dust at the present. I would keep the column temp. as low as possible during the injection in order to promote re-focusing on the column. I even went so far as to use sub-ambient cooling in the oven. This helped quite a bit. I read somewhere about a gizmo that uses pizzo-electric cooling and is attached to the bottom of the injection port inside the oven. This has the distinct advantage of not requiring CO2 or LN2. Good luck!
By Mike Szelewski on Monday, December 11, 2000 - 01:12 pm:
I haven't run a Saturn 2000 LVI, but I do have LVI experience on an Agilent 6890/5973. I'm assuming you're using a PTV type inlet.
Answers to your questions:
1. The goal is NOT to evaporate the solvent while the extract remains in the needle. You always want the sample in the inlet to be the same as the sample in the vial.
2. You do want to inject fast, if there is enough liner volume to hold the liquid, or if you're evaporating the solvent fast enough(see 4 below)otherwise you'll get excess liquid solvent on your column. However, if you inject onto fused silica or glass wool, active analytes may degrade. I've seen that most wools are active to varying degrees.
3. I typically keep the inlet at the solvent boiling point, and the column low enough to trap your earliest eluter. Don't start the column program until all solvent has vented from the inlet and the analytes have been transferred to the column by heating the inlet.
4. Rate of injection, inlet temperature during injection, vent flow and solvent boiling point are all inter-related.
5. I can think of two things that might cause the r.t. shifts. First, excess solvent on the column, and second, not holding the column at a low enough temperature long enough.
6. Sorry, don't know of a book on LVI.
One last point - a common problem new users of PTV experience is loss of early eluters. It's very difficult to evaporate just the solvent, if your earliest eluter b.p. is close to the solvent b.p.
By Ron on Tuesday, December 12, 2000 - 06:56 am:
Mike makes very good points, and I can't disagree with anything that he says. I my experience, some quartz wool or deactivated glass wool will give more surface area to hold analyte while solvent is evaporated, but there is a risk of loss of analytes. I usually do see some loss of the most reactive analytes, but in may cases the larger volume injected does increase the sensitivity, just not to the theoretical prediction.
I have had better luck holding the inlet slightly below the boiling point of the solvent. The solvent still evaporates rapidly, but less appears to transfer to the column, based on less solvent tail.
If you are not evaporating the solvent adequately you will get a large solvent peak with a long tail. If this happens raise the split flow or evaporation time before increasing the inlet temperature during injection.
Early eluters are typically a problem. Using a mg or so of a sorbent may help, but you will probably lose active components.
As to a book, there has to be an understanding of the samples and the chemistry involved to successfully analyze samples using LVI. It is not a technique where you can just enter cookbook recipes and expect to get good results, as I'm sure you've discovered from your experience.
By John Beland on Tuesday, December 12, 2000 - 10:32 am:
OK .. The low initial oven temp makes sense, and the low initial injector temp. But what about injection rate? How do you figure out how fast the solvent is evaporating? If you inject too slowly, your analytes end up concentrated in the tip of the needle, on a droplet on the end of the needle (which gets wiped off on the septum when you remove the needle), or as a solid on the tip of the needle (same problem). If you inject faster w/o a frit or wool, the droplets from the needle fall past the column or directly into the column, neither of which seems desirable. And what about the occasional droplet that adheres to the end of the needle? Just assume it's small relative to total volume injected and go?
FWIW, my seat-of-the-pants technique is to rapidly expel the desired injection amount of solvent onto the SS apron on the front of the hood, and time how long that takes to evaporate. I then inject my total volume over a time ~30% less than the evaporation time. Anyone have a better method for guess-timating injection rates?
I guess I'm trying to conceptualize what's going on and what we're trying to do in the LVI injector. Sorry if all these questions sound dumb.
By Mike Szelewski on Tuesday, December 12, 2000 - 02:01 pm:
A lot of good questions. You're right, don't inject too slowly. How slow is too slow - trial and error. Or you could call it "running experiments". I find best all around results injecting as fast as possible, and adjusting other parameters. For instance, instead of injecting 50 uL fast in one shot, and dribbling down the liner, try 2 shots of 25 uL each. If you use a multi-baffle liner, the droplets won't fall directly into the column.
Ron's correct - raise split flow or vent time before raising inlet temperature during injection. Less chance of early eluter loss.
Ron's also correct in that there is no cookbook.
Solvent type, injection volume & speed, inlet temp, vent flow & time are all related as well as range of analytes.
Also, don't forget to contact the applications chemists at the manufacturer. They'll have a very good understanding of the optimization routine for their particular inlet.
By Ron on Wednesday, December 13, 2000 - 10:59 am:
Mike hinted at something above that is one of the major reasons there are no simple cookbook recipes for LVI. All manufacturers LVIs behave a little differently, so a method developed on one inlet will more than likely have to be modified to work properly on a different inlet.
One reason I like to use quartz wool in the inlet in spite of the potential activity problems is that injection speed becomes less critical. The larger surface area allows more rapid injections than is possible than in an unpacked inlet.
Mike is correct about evaporation being easier from repeated smaller injections than one large injection. The time between injections will allow most of the solvent from the prior injection to be swept from the inlet. Just remember that you are making more injections and septum life will be shorter.
The questions have been good, and show why LVI is not more common. It can take a lot of time and experiments to optimize the methods.
Posting is currently disabled in this topic. Contact your discussion moderator for more information.