Turn-on fluorescence sensors, as you may know, are the type of sensors which do not show any fluorescence emission peaks in a specific wavelength till they bind with the target analyte and that cause appearing a peak at mentioned wavelength.
All of us are familiar with 3S/m formula, but the question is when you do not have any emission peak for blank, then what formula can be used to measure LOD?
(I do not want to use graph to find LOQ and then convert it into LOD)
I'll be appreciated if you share any papers or references about this topic with me.
Dear Dr. Aurea Andrade-Eiroa and Dr. Farooq Wahab thanks for your answers, would you mind please checking the photo I attached to the question, in our specific wavelength the intensity of blank is less than 10 units (a.u.) and the intensity of analyte is about 450 units, by repeating the scan, the blank's intensity changed from 5 to 10, I was thinking it's just a noise and you mean that we should report and use this low amount for measuring the LOD?
I think, in this case, it's better to calculate the LOD using the calibration data and regression statistics . You can use the y-intercept and standard deviation of the regression: y(lod)=a+3s(x/Y). This value is closer to the real LOD respect to the LOD calculate by the instrumental noise.
Because LoD is an ambiguous term there is no one correct way to measure it. The question you first have to answer is - what will you use this LoD for?
If you just want to find LoD just to characterize you analysis method and it wont play an important roll later in the use of the method I suggest you use the ICH proposed method where LoD=3.3*S/slope. S is standard deviation of the residuals. This will give you a good conservative LoD (this is based on my experience with LC-MS). LoD can vary between days/experiments so measuring it on more than one day is a good idea.
If you need the LoD to really show you if some sample contains the analyte or not then determine the average value and standard deviation of blank samples. Then calculate decision limit (Lc) from this data: Lc=average(blank)+1.65*standard devation(blank). (I suggest using at least 10 separately measured blank sample results.) Any signal above Lc is at least 95% times the analyte and not noise. (NB! I suggest repeating sample measurements - if the concentration of the sample is at or close to Lc it is possible (even likely) to get a false negative results.)
If LoD is an important parameter for data interpretation and for characterizing your analysis method then I suggest you look at ISO 11843-2:2000 or IUPAC. They provide you with approaches to estimate decision limit and detection capability. Although estimating these will be more work, the result will be statistically correct and will give more confidence to your interpretations.
I support Hanno...start from the golden rules....
http://goldbook.iupac.org/Graphs/L03540.3.map.html
and clarify your goals in term of confidence levels of your mesures
Thanks for asking. The other responders have given you chapter and verse from the analytical chemistry viewpoint, and I don't disagree with them. I would only add that if you're not seeing any apparent background emission in your blank solution you probably need to compare it to a more dilute solution of your fluorescent indicator + the metal ion. You're probably working at low gain/narrow slits to avoid saturating with the +metal solution, so you don't see the background. Although the reviewer probably wants to know the classical detection limit (which is usually determined by the affinity), what you really need to know is the relative level and variation of the different noise sources, chemical and electronic: background fluorescence of the solution, shot noise of the detector, electronic noise of the amplifier, etc. If you're analyzing natural waters or biological systems, our experience has been that the variability in background fluorescence is frequently an issue.
I was pleased to see that you provided such a quick response to the question I sent in.. I'm really appreciate for your answer Dr.Thomspson, and thank you again Dr.Aurea Andrade-Eiroa for your complete explanations and thank you so much Dr.Chiavarini, Dr.Troisi and Hanno Evard for your time and answers.
- Badges
- Science topic
- Similar topics
- Chemistry
- Biochemistry