Discover how the Thermo Scientific Q Exactive System enables simultaneous qualitative and quantitative analysis of pesticide residues in foodstuffs.
Dr Hans Mol (Senior Scientist, RIKILT Wageningen UR) describes the favored combination of full scan and vDIA (variable Data Independent Acquisition) for targeted and non-targeted approaches and how you can obtain efficient data analysis. A summary of the Q&A session is provided below. To watch the webinar on-demand, click here.
Q: What is the difference in sensitivity between the Q-Exactive and a triple quadrupole instrument?
A: That depends on the compound and the triple quadrupole instrument that you are using. The newest, high-end triple quadrupole instruments would be more sensitive than the Q-Exactive although it is difficult to give an exact figure because you are comparing the fully dedicated, optimized acquisition of a triple quad with a full scan measurement using generic settings, which by definition will slightly compromise sensitivity. The important question here is whether the level of sensitivity is fit for purpose to which the answer is, for this application, yes.
Q: Where I can buy/get a matrix or blank of fruits or vegetables free of pesticides for posterior uses in the validation?
A: Usually we take a sample that we have analyzed before, but if you don’t have that then the best way to go forward would be to buy an organic sample and usually (except for some pesticides which are allowed in organic production) these samples will be free from pesticides.
Q: Have you done any pesticide determination in dietary supplements using this technique? Are dietary supplements considered food in the EU?
A: We haven’t tested any dietary supplements so far for pesticides, although we have tested them for other contaminants. Dietary supplements are considered food under European law.
Q: What is the uncertainty of your validation method, and how do you measure it?
A: What I described in this presentation is the initial validation. What typically happens after that, once you implement the method and you start using it, you include spike samples again and with that you gain more data on the recoveries. Based on the average recoveries you obtain reproducibility and from this we calculate the measurement uncertainty. With this application the standard measurement uncertainty would be between 30-50%.
Q: With the acquisition method you showed in your final method, you had a quite long duty cycle time (978 ms), is that compatible with UHPLC?
A: With the chromatography that we used, our peaks were approximately 7-8 seconds in width, which is relatively low but we were still able to get good quantitative data. If you were to go for a really fast UHPLC method with narrower peaks (~5 seconds) then you may want to increase the overall scan speed. One way to do that would be to reduce the resolution of one or more of the CTIA events from 35,000 to 17,500.
Q: How much time does the data processing take? And how much time does it take you to go through the screening results?
A: With the software that we use here, data processing of one data file with 170 compounds in the quantitative part takes 15-20 seconds. Going through the screening results depends on how many hits you get but it typically takes 30-60 seconds per sample so for a sequence of 30 samples you would be finished in 20-30 minutes.
Q: What about pesticides in feed (a major challenge)? Does this give fewer possibilities for confirmation?
A: With the resolving power we used here, cell activity is not the main issue when testing for pesticides in feed. What could happen is that due to the more complex matrix, you introduce quite a high amount of matrix which may lead to additional suppression and this may reduce your sensitivity.
Q: In the screening part you gave an example, terbutylazine, which was detected and seemed a true positive but at a very low level. In my routine analysis I am not interested in such low levels so I don’t want to waste time on re-running the extract with calibrants only to confirm that the level is far below anything I would report. Does this happen often and how are you dealing with that?
A: This does happen with some compounds and I agree, you don’t want to waste any time on that, so for those compounds we include an additional solvent calibration standard so we can use the response as an additional filter. Any response that is well below the 10 ppb would be discarded.
Q: There are a lot of pesticides that are never or seldom detected. Did you find any additional pesticides using the screening compared to what you would have found when running a triple quad method?
A: In 2013, together with our colleagues from the Dutch Food and Consumer Safety Authority, we compared a run on a triple quad to the method described here. They were using a triple quad method targeting well over 200 pesticides and we re-ran 180 of their extracts in full scan mode. These samples were a wide variety of food matrices including exotic fruits and samples from import control originating from all over the world. We found 17 pesticides exceeding the 10 ppb that they did not report. With such a wide variety of samples it is difficult to be comprehensive for pesticides with the triple quad so the full scan option is a real advantage.
Q: Do you really need to validate your screening method for every individual pesticide? What if your detection limit is above 10 ppb, is the screening still useful?
A: It depends on the purpose of the additional screening. In principle we are able to screen for many hundreds of pesticides but without validation you cannot guarantee that a pesticide was not present even when obtaining a negative result. You can only claim the absence of a pesticide at a specified level if it is successfully validated but these methods can still be used for picking up some pesticides. It all depends on communication with the customer.
Q: The cycle time mentioned in your presentation is only in positive mode. Does it mean you'll double the time by working positive/negative switching? If yes, how you will manage taking into account the non-targeted analysis strategy?
A: Yes, it would approximately double the time if you ran the cycle time in positive and negative modes. It would require two injections and although this takes a longer time it gives you the opportunity to optimize the LC eluent for your negative pesticides.