****************************************************************************** From: bruce_f@nelson.planet.org.nz (Bruce Fraser) Subject: Ion traps vs quadrupoles Date: Wed, 06 Dec 1995 21:40:14 +1200 Organization: PlaNet Community Networking (Canterbury, New Zealand) Can anyone help me? I am evaluating ion traps (specifically the Saturn 4) vs quadrupole GC-MS systems and the following issues have come up It seems clear that the Saturn is quite a bit more sensitive than the quads however I have have been told... * The performance of the ion trap goes down markedly when the source gets contaminated. Ion traps are reputedly much more sensitive to this than quadrupole systems * They still give non classical spectra => problems when using commercial libraries such as NIST/EPA/NIH to search. Also less robust in litigation situations * Spectrum varies with the amount of the component present => problems with quantitation. Also quantitation of a small peak chromatographically partly (or not) resolved from an adjacent large peak is not good. * The ion trap has trouble coping with the amounts of water introduced into the system when doing Purge and Trap sampling. Can anyone comment on the truth or otherwise of these assertions and relate their experiences either with this system or with having a similar evaluation process. Also I am aware of a paper in Anal Chem 1989, 61, 2500 - 2503 by Fales et al on Intermolecular Processes in the Ion Trap Mass Spectrometer. Are there more up to date papers in a similar vein on ion traps, especially the Varian Saturn 4 system. Any references would be appreciated. Regards Bruce Fraser 19 Ngatiawa St, Nelson New Zealand ph: +64 3 545 6016 fax: +64 3 545 7017 Email: bruce_f@nelson.planet.co.nz ****************************************************************************** From: Bob Hendry Subject: Re: Ion traps vs quadrupoles Date: 6 Dec 1995 22:57:15 GMT Organization: Georgia Tech Research Institute We have a Saturn which during the first 3 months of doing air monitoring did not need to be cleaned and performance was wonderful. Although when doing injections of a sludge like matter the traped needed to be cleaned every week to keep performance acceptable. But the cleaning of the trap is very simple compared to the quads in our lab. Our lab has several quads and one Saturn, but the Saturn does most of the work because of its sensitivity and flexablity. > >* They still give non classical spectra => problems when using >commercial libraries such as NIST/EPA/NIH to search. Also less robust >in litigation situations I have run into some problems with library matching, the correct compound not in the first three hits, but the correct hit is usally inthe top ten. > >* Spectrum varies with the amount of the component present => >problems with quantitation. Also quantitation of a small peak >chromatographically partly (or not) resolved from an adjacent large >peak is not good. > The spectrum variation I have seen is when the trap is overloaded with the compound of interest. I have not had problems with the software quanitation of a peak from coeluting peaks. >* The ion trap has trouble coping with the amounts of water >introduced into the system when doing Purge and Trap sampling. > I have not done Purge and Trap on the Saturn, but have had problems with water in air monitoring with the VTAT injector (none that could not be solved). >Can anyone comment on the truth or otherwise of these assertions and >relate their experiences either with this system or with having a >similar evaluation process. Overall I love the job the Saturn ion trap has done in our lab. -- Bob Hendry Georgia Tech Research Institute Atlanta, Ga 30332-0820 Phone 404-894-3825 FAX 404-853-0113 Internet bob.hendry@gtri.gatech.edu ****************************************************************************** From: chris stoneking Subject: Re: Ion traps vs quadrupoles Date: Tue, 12 Dec 1995 12:02:54 -0800 Organization: Georgia Institute of Technology What Bob doesn't mention is that most of the compounds he is hunting for are easily identifiable, i.e. they have very distinct spectra. I believe Bob is in a fairly typical analytical setup and his comments are valid. Now, if you want to know about problems, lets talk about the Saturn's software. It sucks. Big time. They wrote the software focusing on the EPA guidelines for air toxics and such. Because of that there are some major wholes in the abilities of the software. Furthermore, whoever wrote it wasn't a chromatographer and probably wouldn't know one if it snuck up on him and bit him on the ass. These comments, by the way, go equally for Finnigan's Magnum software, which is virtually identical to the Saturn. I know of someone else who did have / is having auto-CI induced by water in the sample. He has a Varian Saturn with one of their VTAT systems on top. (I don't recomend that baby either) It comes down to what you are going to use it for. If you have a straight forward analytical lab setting where you either know ahead of time what you are looking for or are willing to live with the library matches, you probably won't have any problems with the Saturn. If you are in a research field where you will eventually have to figure out for yourself what that odd peak is because you are certain protein fragments don't show up in process water from settling pond #4, then you might think twice about the ion trap systems. ****************************************************************************** From: Jeff Corkern <74030.52@CompuServe.COM> Subject: Re: Ion traps vs quadrupoles Date: 15 Dec 1995 02:05:02 GMT Organization: 74030.52@compuserve.com I might add that it would be nice to know just how much sensitivity you need. I have 2 Hewlett-Packard 5971A's I use for environmental analytical work and they are excellent machines. The support is expensive but is (normally) also excellent. HP's software is good but documentation is poor. What you are going to do with your MS is a critical question. Mass specs are MUCH more sensitive than most instruments. I worked in a GC/MS lab in a university setting once and most of the samples I shot had to diluted about 1/1000. Scanning 40-550 amu in one second, an HP 5971A can get good spectra on 10 ng of compound. From your question, I suspect customer support might be more important than absolute sensitivity. Even routine gadgets like a 5971A can see very small amounts. Customer support should be part of your purchase decision. Good luck. -- Jeff Corkern ****************************************************************************** From: Bob MacPherson Subject: Re: Ion traps vs quadrupoles Date: Thu, 14 Dec 1995 23:03:00 -0800 Organization: Laboratory Equipment Exchange chris stoneking wrote: > > What Bob doesn't mention is that most of the compounds he is hunting for > are easily identifiable, i.e. they have very distinct spectra. I believe > Bob is in a fairly typical analytical setup and his comments are valid. > > Now, if you want to know about problems, lets talk about the Saturn's software. > > These comments, by the way, go equally for Finnigan's Magnum software, > which is virtually identical to the Saturn. I would be curious to know which software you feel is best for MS. I think all that is left is HP and VG. I've seen both of those, and I agree with your general assessment on the Finnigan and Varian S/W. VG has done some neat things, and I think HP has some newer software that I haven't seen. > It comes down to what you are going to use it for. If you have a straight > forward analytical lab setting where you either know ahead of time what > you are looking for or are willing to live with the library matches, you > probably won't have any problems with the Saturn. If you are in a > research field where you will eventually have to figure out for yourself > what that odd peak is because you are certain protein fragments don't show > up in process water from settling pond #4, then you might think twice > about the ion trap systems. How would you compare a Saturn to a HP MSD in a routine setting (know what you are looking for)? -- _________________________________________________________________________ Bob MacPherson Laboratory Equipment Exchange http://www.magic.mb.ca/~econolab Your Internet Magazine for Used econolab@magic.mb.ca Scientific Equipment four star rated by the mckinley _________________________________________________________________________ ****************************************************************************** From: mikemcd@axionet.com (Mike McDonell) Subject: Re: Ion traps vs quadrupoles Date: Fri, 15 Dec 95 23:56:06 PST Organization: Axion internet >How would you compare a Saturn to a HP MSD in a routine setting (know what you are looking for)? >--The comparison of traps vs quads has so far not mentioned the fundamental difference in ion source characteristics of either device. These differences are obvious to me but may not be to everyone. So for those interested; Quads: External ion source, continuous ion beam, generally filtered to 1 amu bandpass and sensitivity limited in full scan primarily due to duty cycle, hence the need for selected ion monitoring ( if you know what you are looking for ). Classical Traps ( Saturn and Magnum ) Internally ionized traps, electrons are gated. GC effluent is continuously dumped into the trapping volume. Collisions between ion and neutral species concentration dependant. Changes in trap spectra over GC peaks may be caused by self protenation or space charge effects. GCQ ( Finnigan ) External ion source, ion beam is gated. During ion accumulation, the ion beam passes through the trap and an extraction waveform will divert a specific mass into a stable trajectory. This waveform is scanned from high to low. Notches are left at masses not to be accumulated. ( Selected Ion Accumulation). After filling the trap, the ion source is gated off, and you are left with only ions and the bath gas in the trapping volume Hope this clarifies things a bit ****************************************************************************** From: mrheadmstr@aol.com (MrHeadmstr) Subject: Re: Ion traps vs quadrupoles Date: 16 Dec 1995 12:25:53 -0500 Organization: America Online, Inc. (1-800-827-6364) In article <4a4el0$s84@acmey.gatech.edu>, bruce_f@nelson.planet.org.nz (Bruce Fraser) writes: >Can anyone help me? I am evaluating ion traps (specifically the Saturn >4) vs quadrupole GC-MS systems and the following issues have come up I recently purchased a system and went thru the same Q/A process as you are doing now, and wound up buying a Finnigan GCQ. Here's some of what went into my decision making process. (And this is a former HP field service guy talking!) 1) Mass Range. The GCQ ranges from 10-1000amu. The HP MSD is up to 650. OK for standard EPA work, not good for work with more variety. 2) Capability. The GCQ offers EI+, EI-, CI+ and CI-. The Saturn also offers CI but it is a $20K option! 3) Data System. The GCQ is written in C++ to run on WindowsNT, a full 32 bit OS with a good reputation for stability. It came with a Gateway2000 P5-75, with 1.3GB hard drive, 16MB RAM, a Quad speed CD-ROM. 4) Library. Either NIST or Wiley on CD-ROM. 5) GC flexibility. Can get it with either the new Finnigan GC or the HP 5990. 6) EPC. Comes with the HP electronic pressure control standard. Split/Splitless capillary is standard also. 7) MS/MS. THE reason why people buy an ion trap instead of quadropole, isn't it? 8) Price. I paid $72,000 for mine with capillary, EPC, MS/MS, EI/CI/NCI, NIST library. I couldn't get the vacuum interlock (a STRONGLY recommended >$6000 option) due to budget constraints. For the same price, I could have a HP MSD without any options, or a Saturn without CI/NCI (and with an unsat software system). IMO, right now, the GCQ offers WAY more bang for the buck than any other GC/MS on the market. Being in New Zealand, service has to be more of a consideration for you than for me, and in that regard HP is clearly superior. If you have a fairly local Finnegan service rep of course, then service might not be a worry for you. Good Luck. Wishing I could come visit your beautiful country someday! ****************************************************************************** Date: 17 Dec 95 15:35:20 EST From: David Sparkman <75530.3472@compuserve.com> Subject: Re Ion Traps I have to comment on the recent questions regarding quadrupole ion trap mass spectrometers. An ion trap GC/MS requires a more experienced operator than a transmission quad. When working with unknowns, strange results are possible due to ion-molecule reactions that may take place between odd-electron ion fragments produced by polar analytes and the analyte molecules. On of the best examples of this is with 2-octanone. The mass spectrum obtained using an ion trap can have one predominate peak at m/z 129 (M+1). This is because the odd-electron ion with an m/z value of 58 is a very good proton donor and will protonate the 2-octanone molecules present in the trap as the m/z 58 ions are formed and stored. All of the rest of ions normally formed are present, but their abundance pales by comparison to the M+1 ions that are formed. The presence of peaks representing the M+1 ion can be avoided by not storing ions below a value that will exclude the fragment at m/z 58. However, you can still see the presence of this important odd-electron ion peak in the spectrum you obtain. The ion trap does not take all the data for a single spectrum in a single scan. For those instruments where ions are formed in the trap, the scan is divided into four different segments. A separate ionization takes place to produce ions to be analyzed in each segment, However, all the ions form m/z 20 to m/z 650 are formed AND STORED during each ionization. If you want to avoid the presence of protonated molecules in spectra of polar compounds, set the RF to store all ions between m/z 20 and m/z 650 in the first scan segment. This is the default storage voltage, and will allow the detection of all ions present between m/z 20 and 99. This segment is used to record data on ions between m/z 20 and 99. Set the storage voltage for the other three scan segments to store only ions between m/z 90 and m/z 650. Most of the odd-electron ions that are very good protonating agents formed by the rearrangement fragmentation are below m/z 100. This will reduce the sensitivity in these segments and may require an adjustment to the ionization time. Use a mixture of ketones and methyl esters of fatty acids to set the conditions that will give "classical" spectra. Then use these conditions for the analysis of your unknowns. The overall sensitivity of the ion trap will be reduced, but it is still as sensitive as or more so than a transmission quadrupole. With the ion trap, you will have the advantages of low pressure CI and MS/MS to get even more qualitative information. You should try to avoid the presence of water in trap during the time some analytes are being looked at. With respect to EPA purge and trap, a lot has been written on this, and EPA accepts the data you can obtain from the trap. Regards; David Sparkman ****************************************************************************** From: nonsuch@netcom.com (David Mcnerney) Subject: Re: Ion traps vs quadrupoles Organization: NETCOM On-line Communication Services (408 261-4700 guest) Date: Mon, 18 Dec 1995 06:15:25 GMT I hope that this message posts as planned, as it is my first post to a newsgroup. I have been using a Saturn ion trap GC/MS for the past two years, mostly to analyze extracts of environmental samples for semivolatile compounds, but also with purge and trap. The company I work for owns two Saturns. I'll atttempt to answer your questions... keep in mind that I have no experience with other types of mass specs such as the HP MSD, etc. The ion trap is actually quite robust with regard to analyzer contamination. I inject some extremely dirty extracts into the machine and have never experienced any serious degradation of performance over time as a result. The analyzer is easy to clean, and we do so regularly mainly as a preventive measure. I'm sure that spectra produced by the Saturn do not always match classical EI spectra very well, because of the way the Saturn uses scan segments to mimic the tuning of other mass specs. Nevertheless, for many compounds I am able to get a very good match to the NIST library. I don't usually have to identify unknowns, so I haven't looked into this problem too much. For some compounds, under some circumstances, the Saturn ion trap will produce concentration-dependent spectra. I have observed this with the phthalate esters in particular. I think that there are ways to minimize the problem, however. The ion trap is extremely sensitive, but its sensitivity varies inversely to the level of chemical background. If two compounds coelute and the amount of one is extremely large while the amount of another is very small, the EICP peak for the small compound may be affected. This has to do with the way that the Saturn varies the time that it opens its electron gate (ionization time) in order to maintain an optimum number of ions in the trap. If you are only interested in the small peak, there are ways to reject the ions generated by the other compound so as to enhance your sensitivity and accuracy of quantitation. Your Varian salesperson will no doubt be happy to explain those, since they cost extra. WRT to the idea that the ion trap is adversely affected by the water in a purge and trap system - I have never heard of this before. It is true that the ion trap is extremely sensitive to the air/water background and care must be taken to keep that background as low as possible, but we used a purge and trap system with one of our Saturns for months with excellent results. I always assumed that the trap in the purge & trap didn't hold on to the water vapor from the purge. As far as differences in sensitivity and performance between the ion trap and the quadrapole, everyone seems to think that the ion trap is quite a bit more sensitive. On the other hand, much of that difference might be offset by starting with more sample, i.e. 25 ml purge or extracting 100 grams of dirt instead of 20, etc. Please email me if you want more information, or if you want to hear about shortcomings of the Varian hardware unrelated to the subjects above. Dave ****************************************************************************** From: chris stoneking Subject: Re: Ion traps vs quadrupoles Date: 19 Dec 1995 17:48:06 GMT Organization: Georgia Institute of Technology Bob MacPherson wrote: > > I would be curious to know which software you feel is best for MS. I > think all that is left is HP and VG. I've seen both of those, and I agree > with your general assessment on the Finnigan and Varian S/W. VG has done > some neat things, and I think HP has some newer software that I haven't > seen. > I don't truly know what is _best_. All the ones I have worked with have some major shortcomings. So far, HP is a little better than the rest, but there are a LOT of instruments I have not worked with out there. > > How would you compare a Saturn to a HP MSD in a routine setting (know what you are looking for)? We do ambient air, looking for ozone precursors and the like. A lot of the time, this means packing up my equipment and taking it out to the field with me. Last summer I spent 8 weeks in a cow pasture outside of Nashville, TN. There are no routine settings for us. I tend to think that side by side in a lab, the Saturn (or Finnigan) systems will be more sensitive and just as well behaved as HP's stuff. And the innards of the ion trap are easier to clean that any quadrapole! Hope that helps, Chris Stoneking chris@aql.eas.gatech.edu ****************************************************************************** Response to similar question in December 1997 ****************************************************************************** From: "Ludwig Gruber" Date: Wed, 17 Dec 1997 13:38:50 +0100 Subject: Re: ion trap vs. quadrapole Organization: Fraunhofer Gesellschaft There is a good compilation about this subject in Analytical Consumer, Vol 6 No. 5 from February 1996. In our experience (we own two ion traps and a numer of quadrupoles) the problem of classic ion traps (e.g. Finnigan Magnum) is the mechanism of the Automatic Gain Control if you have a coelution of peaks. In this case you have troubles with analyzing and quantifiing the minor compound. Regards Ludwig Gruber ****************************************************************************** From: dond@mhv.net Date: Fri, 02 Jan 1998 22:11:41 GMT Subject: Re: ion trap vs. quadrapole Organization: MHVNet On 17 Dec 1997 08:38:13 -0500, "Ludwig Gruber" wrote: }There is a good compilation about this subject in Analytical Consumer, Vol 6 No. 5 from February 1996. } }In our experience (we own two ion traps and a numer of quadrupoles) the problem of classic ion traps }(e.g. Finnigan Magnum) is the mechanism of the Automatic Gain Control if you have a coelution of peaks. }In this case you have troubles with analyzing and quantifiing the minor compound. } }Regards } }Ludwig Gruber Dr. Gruber is exactly right on this. We use both quadrupole systems and an ion trap system (the Finnigan GCQ) in our laboratory for always complex and dirty systems (petroleum and related samples). The ion trap systems have a reputation in our industry of working very well for clean samples at low gc temp (drinking water analysis, etc.), but poorly for complicated dirty systems. The main problem is that there is always tons of stuff eluting (solvent, column bleed, non-analyte-matrix species) that the when the AGC does a "test" ionization to figure out how many ions are present, it always sees a lot, and thus automatically turns its ionization time (sensitivity) down. This means that your detection limits move up and down un-predictably depending on how much is co-eluting in any given sample and you can never guarantee the analyte is not present at any given concentration. The in-trap ionization is also problematic for heavy dirty samples, since for EI, 100,000 neutrals must exit the trap for every ion, which means a lot ends up where you don't want it. The GCQ system mitigates these two problems to some extent, since the ion source is external to the trap, the trap only gets loaded with ions (and background neutrals). It is also possible to use the ms/ms waveform excitation to only count the (limited) analyte mass(es) in the AGC step if you don't need full range spectra. Other manufacturers say they can do this last trick with internal-trap-ionization as well, but I can't speak for how well it works. Other manufacturers say they can treat the trap surfaces with coatings and/or heat to minimize the adverse effects of letting your whole sample (plus column bleed) directly into your analyzer. Hope this helps. Dr. Don Drinkwater dond@mhv.net drinkde@texaco.com drinkde@texaco.com ****************************************************************************** THIRD QUESTION/ANSWER ****************************************************************************** ****************************************************************************** From: "bdguest" Subject: GC/MS Instrument Evaluations Date: 27 Jan 1998 20:59:39 GMT Organization: Insync Internet Services [This is forwarded from sci.chem.analytical. DB] I'm considering purchasing a GC/MS. What do you know about the reliability and sensitivity of the Shimadzu, Finnigan, and Hewlett-Packard instruments? Would appreciate any feedback. I would be buying a new instrument, so only supply info on the latest models please. Dr. B. ****************************************************************************** From: Vladimir Elias Date: Wed, 28 Jan 1998 15:58:23 -0800 (PST) Subject: Re: GC/MS Instrument Evaluation Organization: * Response to: >I'm considerin purchasing a GC/MS. What do you know about the >reliability and sensitivity of the Shimadzu, Finnigan, and Hewlett-Packard >instruments? Would appreciate any feedback. I would be buying a new instrument, so >only supply info on the latest models please. >Dr. B. I don't know about Schimadzu and Finnigan but the new HP 5973 MSD is a powerful and impressive machine. I've work on the previous models (5987, 5970, 5971, 5972) and can tell you that this new model is superior in many aspects. I've been using the HP 5973MSD for 13 months and I'm impressed with its performance (so far!!!). The results are reproducible (even at trace levels as pg), the source is easy and quick to be cleaned (2 hours at most) and don't so many cleaning as the previous models. Cleaning is usually required prior injections of more than 200 silylated samples. The system is also very friendly allowing first time users not so skilled operators to maintain the machine. A CD-ROM video shows step-by-step what needs to be done. An advantage in educational institutions always where students are always coming, learning and going. In my opinion the only disadvantage is when we do high temperature GCMS. As I mention in my recent papers (JMS and HRC), in this system we can just see masses up to 800 uma. As during HT-GCMS many compounds exceed the 1000uma this the limitation of 800uma s not acceptable. But I think HP will have to change this because HTGCMS is growing pretty fast. VOE ******************************************************* Vladimir O. Elias www.orst.edu/~eliasv Oregon State University College of Oceanic and Atmospheric Sciences 104 Ocean Admin Bldg Tel: 541-737-2741 Fax: 541-737-2064 *******************************************************