Agilent 6490 LCMSMS system

[Ken]

Hi there. I hope someone can advise me on the mentioned Agilent 6490 LCMS/MS instrument.

I want to know more about this instrument before I make any decision to purchase. We're mainly dealing with pesticides but we are also going to look out for natural product researches. Hence, if we're just using a TQ, it will not be able to help us with our natural products and metID research works.

We already have TOF based instruments and currently, we want to evaluate this new instrument 6490 from Agilent.

Is there any real life issues encountered when using this instrument? In our country, there is not a single 6490 yet thus we want to be sure so that we won't be the 'guinea pig'.

I know that:
i) Mass hunter software is very resources hungry and tends to slow down data acquisition. If there are many samples being queued up, the computer will slow down and this might cause error in data collection. Is that true?

ii) How about the instrument robustness? It was claimed that the instrument is highly robust but somehow, it was reportedly to be less robust with real matrices samples.

iii) System tends to get dirty fast ie ion source

iv) others

Can anyone advise?

For your info, we currently have QTrap 5500, 4000QT, 3200QT, QStar Elite, Orbitrap, and Waters QTof.

We are planning to use the budget to get another 3200QT but just want to know how is the 6490 performed compared to the rests in our research works ~

Especially any real issues encountered ~ ^^

Thanks and hear from you soon

[zokitano]

Well, you have a nice collection of different mass spectrometers with diffrerent abilities and pros and cons. Starting from triple quadrupole-linear ion trap to TOF and Orbitrap.

Have you tried to do your natural product and metabolite research on those "beasts" that you currenly own. I havent used the Agilent 6490 system but as I can read from the web it is a triple quadrupole MS, in which mode you can also run your samples on AB&Sciex QTraps that you own.

Still if you need evaluation of the performace of the Agilent MS system, why don't you send your "difficult" samples to analyze to Agilent and ask for demo, and then compare the results from their system with the results that you can obtain on the different MSs that you currenly own?

It would be a nice comparison for which I would like also to know the outcome

Have fun with your powerful toys

And good luck!

[Jen]

Hi, Have you purchase Agilent 6490 yet? Do you like it? You have several LC/MS/MS from different vendor, Can I ask your opinion about API4000 and agilent 6430? Which is better? Is 6430 software easiler to use than that of API4000? which instrument is more robustness? which intrument ion source get dirty faster?
Thanks

[Ken]

Hi Jen, no, we decided not to buy the 6490, we opt for the 3200QT.

The reason is simple:-

We sent our samples to them : Sciex, Agilent, Waters, Thermo

Instrument tested by them: AB Sciex - 3200QT, Agilent - 6490, Waters - Xevo, Thermo - Vantage

The best results compared to our own data:- AB Sciex 3200QT, followed closely by Waters Xevo, Thermo Vantage and finally Agilent.

Tested criteria: Accuracy ( as compared to our own data ), repeatability, reproducibility, LOD and LOQ

Though Agilent 6490 has a higher LOD and LOQ, but their accuracy, reproducibility and repeatability under low concentration is extremely poor - in fact, the results are completely different that our own data. In short, we concluded that they have a sensitive instrument but very poor accuracy ~ Xevo is good with good accuracy and reproducibility and repeatability but their result are not as accurate as the from the 3200QT. Vantage is ok but somehow they told me that it's not advisable to run for natural products. So, is a no go.

In anyway, the criteria to purchase this time is:
i) Targeted pesticide samples
ii) Natural products
ii) Budget

So, the instrument that met our budget and criteria/expectation this round is (still): 3200QTrap

Thus now, we have another unit of 3200QT which is solely used for this purposes ~ the older 3200QT is used for other research works :p

As for your questions, about 6490 and API4000, we dont have API4000 but 4000QT ( of course, it's the same except for the trap function in the QT ).
We don't have 6490 but from the data obtained from the demo that we did, we can generally said that both are about the same in term of sensitivity ~ as for robustness, we don't have any Agilents but I am very happy with our Sciex as it has proved to be very robust even with dirty matrices ( for example, I only need to clean our 4000QT ion source and Q0 about 3-4 times a year only with very dirty matrices ie biological samples ( plasma, blood, urine samples - no SPE cleanup ^^ ).
As for software, we are pretty happy too ( it's very subjective to the users - depends on how well verse you are with them - maybe need more practice ) - for example again, I used Cliquid ( from Sciex ) which is pretty much like Tracefinder ( Thermo ) and supposed to be very easy to use and interactive (which it was), yet at the end, I still prefer the original software itself as I have more control and I can do what I want ^^

So, at the end, I think you may try to do demo samples request and let all the vendor test and report on your samples and proceed from there onwards yea ^^

Zokitano: Yes, thanks I am having fun with our toys here - and hope to get more toys soon to play

[yangz00g]

To be honest, I don't quite understand why use Qtrap to do simple QQQ work. You may save more money by buying a API3000 QQQ. I have seen many labs (even in my own company) having a QTap, but never or seldom use the trap function. I also have seen many are highly excited after coming back from the well-designed AB Sciex QTrap training, but only figure out it is not easy to apply some theories into practices.

For your purpose, API4000 QQQ may be a better option within your budget (BTW, may lab have an API5000 QQQ for pesticide analysis).

[Ken]

Hi Yang, thanks for your concern ^^

We are not only doing simple QQQ works but also on unknown screening and fingerprinting works with the trap; hence the decision ~
I must say that you're right that it's not so easy to apply those theories but I am fortunate to say that we are now quite well verse and are able to use the trap functionality pretty well for most of our unknown screening and fingerprinting works in natural products and others ^^ of course, those data will also be subjected and correspond back to the data obtained from our TOF and orbitrap for further works etc etc ^^

Thank you again ! ^^

[Jen]

Thank you for all the valuable information.

But I am surprised that you compare 6490 to API 3200Q. Isn't the 6490 the Agilent top LC/MSMS?

[Ken]

Jen: Yes, 6490 is Agilent top LCMS/MS.

We did not set any models to be tested for our samples but Agilent offered the 6490 in their proposal.
So, we have no choice but to compare the results obtained from all the proposed models.

And, as I mentioned earlier, 6490 has the highest sensitivity amongst all the instrument models but their accuracy and reproducibility/repeatability are way off~

[lmh]

I'm a bit puzzled:

"Though Agilent 6490 has a higher LOD and LOQ, but their accuracy, reproducibility and repeatability under low concentration is extremely poor "

How can you have a low LOQ, but poor repeatability at low concentration? (I'll admit personal bias: I've always defined LOQ in terms of s.d. of low-level calibration curve rather than in terms of S:N ratio). Do you mean that using the Agilent instrument, on any given day, mutliple measurements of a low-level standard agree very well (therefore low LOQ), but if you repeat the determination of LOQ on different days or different instruments, the LOQ value varies?

Also, since accuracy is a matter of comparison of sample with calibration curve (LC-MS never measures quantity on an absolute basis), doesn't lack of accuracy actually imply lack of linearity or a problem in fitting the calibration curve? The only other way you can get systematic bias in results is by making up the standard wrongly, or by using a different standard that turns out to have a different purity/error in its make-up.

[yangz00g]

I personally think that overall Agilent 6490 is better than API3200, since its main counterpart is API5000. Roughly the sensitivity of API5000 is 20x that of API3200 in the QQQ mode.

Ken, Instrument performance was one factor in your trial, but who did the analysis is much more important than the instrument itself. If the samples you sent out were pesticides, you hit right on the target, in my opinion, AB Sciex is one of the best, if not the best, on pesticide analysis.

Do not quite understand how you define reproducibility and repeatability (not very often used in analytical community), and how those relate to precision. I second to IMH on the LOQ issue, LOQ should be based on precision (BTW, many European colleagues like to use the S/N for LOD/LOQ).

[Ken]

Hi Yang, LMH, both of you mentioned it correctly and well.

Sorry to make the confusion ~

Well, we defined LOD, LOQ based on european regulations ie S/N in term of s.d, etc etc. Anyhow, I think I will not dwell too much on the criteria/requirements. I think both of you guys are well knowledgeable on it and we (with utmost humility) are also quite well experienced on the regulations and analytical requirements for intrument evaluation, testing, validation etc.

Just want to 'save' some time ~ I apologised if the previous comments are not too complete

And yes, of course Agilent 6490 is much more sensitive than 3200 based QQQ LCMSMS.

We strongly believed one of the main reasons why the results differs so much is also because of the technical capabilities/support of the team which is one of the more important criteria for us too. If they are not able to provide the support well, there is no point in us getting the instruments. And yes, at low concentration, their RSD values were poorer ~ is it the instrument? is it the technical capabilities of the support team? etc etc ~ If they really have good RSD at low concentrations with their specs, than it means that the team did a poor evaluation job ~ so

All vendors get the same standard, sample prepared at the same time from the same source - and transferred into different vials.

[lmh]

Ken, you've put your finger very accurately on something important: it's extremely difficult to avoid testing the demo chemist instead of the instrument! Since having the manufacturer's active support was obviously a particuarly relevant factor for you, I'm sure you were right in your decision. I always think people should draw up their own list of criteria, and carry out their own comparisons and tests (as you did). Recommendations on sites like this are rarely useful; everyone's needs are different. We are lucky that there is a good handful of manufacturers out there who are all very different, each with particular strengths.

Incidentally, I'm always glad I'm not a demo chemist. Not an easy job.

[Jen]

Do you have any problem with API 4000 dwell time? Another competitor suggests their LC/MSMS have faster dwell time and therefore can do more MRM and is faster enough to use UPLC?

[RobKen]

Is there a new ABI 4000? I have seen a spec sheet for ABI 4000 which has interesting numbers on it. Like 200fg reserpine on column giving a S/N 1200, scan speed 24.000 , but still the polarity switching is 700 ms. I was thinking that ABSciex might have using different electronics but in that case 700 ms dwell is still the same. And there is no news about this on ABSciex web site. Anyone had any idea?