Triple Quad Choices.

[matthew]

Hi, sorry for the umpteenth query about purchasing lc/ms/ms equipment in this forum, but I'm not sure how people respond to yahoos digging up old posts here.

I'm a student in an ecology lab, and we intend to quantify leaf metabolites found in various accessions (individuals, lines, strains) of a local plant species. To that end, we are going to purchase a triple quad ms.

I'm impressed with ABI's instruments; there is a good community of instruments on my campus (user groups, etc.) and the QTRAP, in particular, seems very useful (triple quad, LIT for ms^n work). We're budget limited, so we're considering the 3200 QTRAP from ABI.

The other machine that we are considering is the Agilent 6410. It is apparently much more sensitive than the 3200 QTRAP, but I'm not really sure how important that is for the range of 'bioactive' metabolites that I'm interested in (e.g. is a plant going to use a femtogram of material to poison a caterpillar?).

I am writing to ask:

1.) Have you had experience with either of these machines? For example, how do you feel about 'Analyst'? I met with someone here that has a 4000 QTRAP and he mentioned that his machine sometimes freezes up.

2.)
a.) What sorts of levels of detection are you getting with your 3200?
b.) Do you use your 3200 QTRAP just for tandem-ms work?

3.) Eventually we'll work with unknowns. There are apparently different definitions of 'unknown', but after speaking to two different chemists, I got two different pieces of advice:
a.) "You won't need to dig out your nmr notes from ochem if you get a tandem-ms, because the structural information from the tandem-ms will be much more helpful than information from nmr".
b.) "Who said that? That's ridiculous".

Q: If I have to do quantitation in a species with several unknown metabolites, should I get the more sensitive instrument or the more flexible instrument? Is it a moot point: The Agilent rep argued that no one does structural work with the 3200 QTRAP because it is an "ancient machine with a new source".

Thanks for any comments or advice.

[JGK]

If you're considering making a significant purchase like an LC-MS-MS, talk to the sales reps and ask if you can get a demonstration of the instrument with an applications specialist using some of your samples.

Discussing your needs with someone and getting a practical demontration of the instruments capabilities relative to your needs will be better than making the decision "blind".

[gxch43]

I have experience wit h the API instrument, following are just for your informations
1)
ABI's instruments is very good except its analyst software, because the software often freezes up for unknown reasons, you have to restart the computer,even though you need restart the LC/MS/MS;

2)3200 QTRAP is not just for tandem-ms work,but it is really an "ancient machine with a new source",it maybe can not meet your requirements in the future.

3)I agree with the second person, nmr notes are very important and it is not replaceable.

[yangz00g]

Interesting discussions.

We have both API 4000 and 4000 Qtrap, so far we haven't had a chance to explore the advantages of QTrap( we use it as a QQQ). Thinking this way, two MRM will give you the needed confidence to confirm an identity in most cases (>99.9? Of course in rare cases two compounds have the same major MRM such as codeine and hydrocodone, at this point, a third MS may be helpful). So IMO, a QQQ will be enough for your purpose.Since none has superior advantage over the others, I will pick the one with readily available service in your area.

3) depending on how you define "unknowns", I think both people make their points. If the "unknown" is a new compund never found before, NMR is a must. If the "unknown" may be just a "new" compound in this typical sample, and been found elsewhere, MS/MS may give a right identfication (of course you need know how to interpret the MS info.)

Q: quantititaion of unknown metabolites is not viable by LC-MS/MS if there is no pure standards available. In this case, neither sensitive nor flexiable is the issue.

[matthew]

Thanks for all of the comments.

JGK: That's the best advice I've gotten. I've asked one of the vendors if we can run our extract on their instrument (e.g. leaf extracted in etoh, put on ods and n2 evaporation of the methanolic fraction) and there's a great deal of pushback there. The concern was that we wouldn't have time in a visit to optimize pH concerns for ionisation in a matrix as complex as a plant extract. I'm reluctant to visit a vendor for a cookbook demo (I put the casserole in the oven and bam 2 seconds later, it's done), but that's all we've got from that vendor (?).

gxch43:
1) It appears that Analyst really has issues. What happens when you have to restart the computer? If it happens during a batch, do you lose everything??
2) For my project, I assume the 3200 QTRAP will be adequate, but I should provide full disclosure to my boss (lest another student join the lab downstream that needs femtogram sensitivity).
3) That makes sense, certainly.

yangz00g:
I've heard from several chemists that the QTRAP will be the machine of the future, but you're the third person that has mentioned using the machine only as a QQQ.

For publication purposes is an MRM sufficient for identification (given your example of codeine and hydrocodone)? I am primarily interested in quantitation, but again, unknowns will 'demand' quantification too if they are 'ecologically interesting'. Of course, when I use the word 'unknown' the MS vendors ignore my interests in quantification and immediately try to upsize me into a QTOF that doesn't fit my needs (and busts my budget), but what else do you expect from a scientist turned salesperson. I digress.

Your point about quantitation of unknowns is key. Stupid question: when someone chooses an analogous standard, instead of a labeled standard, that is still based on full knowledge of the identity of the compound, I assume(?). Is the whole point to avoid suppression in that case? Is it a cost saving technique?

Thanks again!!

[yangz00g]

Matthews,

IMO, no matter it's a linera or traditional ion trap, quantititaion is the weak point of an ion trap MS comparing to a QQQ, however, you will see improved sensitivity.

For publication purpose, in the codeine and hydrocodone samples, two MRMs (including the quant one) are enough based on the prerequiste those two compounds are well resolved by LC. The QTrap will be very help when additioal identfication is needed and/or for a rapid MS/MS screening (no LC). But you will get lower sensitivity using MS/MS/MS transition for quantitation.

You can use a very similar analogous (same function groups, only one or two difference in CH3) as the IS (we do the same, but as long as we can find isotope-labled IS, we use that), but as you know, ion suppression is the major problem associated with current MS/MS and its cause is not very clear. The analogous may not bahave the same way in the interface, which may bring some troubles.


More to consider: I don't know why, to date, Agilent and ABI still cannot get the amchines reconclied as well as many expect. You still need two persons to install/train LC and MS seperately, and connetion between the two machine is very old fashion (cut the cable, put the copper wire out, stick them into the back of HPLC).

[matthew]

Hi yangz00g,

I see your point about using two fragments in two MRMs to distinguish isobaric compounds... definitely a cool approach.

I forgot to ask you earlier: have you also had trouble with Analyst? Have you had to reboot the machine during a batch? I assume that you don't lose your data, since AB would have trouble selling machines if that were the case.

Also, do you know the scan speed of your API 4000? I think that the Agilent 6410 is supposed to be compared to it, rather than the older 3200 or 3200 QTRAP.

Thanks again!

[JGK]

Matthew,

In my previous job, we purchased an API4000 and were running on Analyst 1.4.1 from 2005. I can't recall any software problems causing the PC to crash and halt data aquisition.

However, our configuration was not "standard" as we used a Waters LC in "Stand Alone Mode" and batch acquisition activation was by a contact closure.

[brewer]

Hi,

I have been using Analyst for the last 7 years, and while it is true that there are some bugs associated with this software, they are not fatal errors.

I have found that as long as you have the recommended 2G RAM, keep your cache clear, and restart the computer regularly (daily or weekly) the stability of the software is rarely an issue. Common issue include a lag in screen display while in autotune or exploring data files under acquisition. I have witnessed shutdown of Analyst during an analytical run on rare occasions, but in each case the instrument continued to acquire data and nothing was lost.

The fact that you're getting pushback on a demo is BS. If they want to sell the instrument, they will extend their visit.

Best of luck!

[yangz00g]

Mattews,

Based on my experience with different machines, anyone freezes or loses communication duirng the run once a while. So as long as it doesn't happen very frequently, I wouldn't bother.

Agilent 6410 is compatible to API 4000 or even 5000. Their scan speeds may be up to 5000 amu/s or even more.

[promochrom]

You may talk to Waters also. I saw many labs using UPLC-MSD combination. When you get more potential suppliers, you have more bargain power.

I know ABI, Agilent, and Waters provide presale sample testing for LC-MSD. If a supplier do not want to provide such service, their application support could be a problem after you have their instrument.

PromoChrom
www.promochrom.com

[matthew]

JGK, yangz00g and brewer:

Thanks, your experience echoes what I've heard here: the software isn't insufferable, but it could be improved.

I visited with Agilent yesterday, and we're leaning in that direction. If the i-s get dotted and the t-s crossed, I believe that we'll end up with the 6410. It seems solid, and the software seems sturdy. I should add that a researcher here raised a issue about Analyst that brewer also hinted at: reviewing data files or trying to quantitate on the workstation while the instrument is running leads to problems. 'MassHunter' seems to be able to accomplish that task, although I should admit that Agilent's user-base is lacking, so I don't know what the criticisms might be.

Wish me luck.

Thanks promochrom; I met with Waters early on in the process and they weren't able to put together a functioning demo when we visited their office (on either of two separate machines... an SQD and a TQD... IN THEIR OFFICE). I think they had changed computers recently, but it wasn't an impressive outing. We would have explored it again, but honestly AB has made this whole thing move... they've been aggressive in their pricing and their user-base here on campus is second to none. I've been tempted to reward their rep's honesty and efforts, but Agilent has a big mo-bio presence here, and they're motivated to establish themselves in this arena (through pricing).

Honestly, if my region's AB and Agilent reps are representative of the rest of their sales force, Waters doesn't stand a chance... unless they keep trademarking things that everyone already knows about or implements. Hey guys, a small particle size, whoa! pthththt. Get a 2.2um column and save 60k.

I digress.

[MS made easy]

You should consider a Thermo LTQ-XL linear ion trap. The linear ion trap can rival the triple quads you mention for quantitation work and will outperform any nominal mass system in full scan MS N for structural work.

[matthew]

Hi 'MS made easy',

Do you work at Thermo or with one of their partners? Your moniker seems to suggest that you have a stake in my purchase.

If not, do you have one of their instruments? Quantitation with a LIT seems hokey.

[MS made easy]

Yes, I work for Thermo. Quantification with a LTQ is not hokey. There are many labs using traps for quantitation that does not require a 350K dollar triple quad. If there are enough ions of interest in the trap relative to the background, the trap does a nice job. There was a presentation at ISSX a couple of years ago comparing traps to triple quads, see below.

But in the end, I agree with JGK, you need to run your sample on the instrument you are going to purchase......before you purchase.



FEASIBILITY STUDY OF QUANTITATIVE LC-MS ASSAYS ON SPHERICAL AND LINEAR ION-TRAP MASS SPECTROMETERS IN COMPARISON TO TRIPLE QUADRUPOLE MASS SPECTROMETERS APPLIED TO BIOANALYTICAL RESEARCH STUDIES Dieter Drexler1*, Petia Shipkova2, Bob Langish2, Joanna Zheng3, Mary Ellen Salyan2, James Smalley3, Timothy Olah3, Mark Sanders3 Bristol-Myers Squibb - Pharmaceutical Research Institute 1Wallingford, CT; 2Hopewell, NJ; 3Lawrenceville, NJ presented at the 7th International ISSX Meeting August 29-September 2, 2004; Vancouver, Canada