I just wanted to thank your entire family! Your brother Greg helps me train harder the last time, and you've helped me study harder than last time for my chemistry exams.
Thank you for the very good explanation, I have learned a lot. Just one quastion: is the response factor (f), that is 1.4 in this example, always constant over the whole concentration range? Meaning, can we assume that is it is 1.4 at 0.1/0.1 M/M, it will also be 1.4 at 0.01/0.01 and 1/1 M/M respectively?
I have a question: Does the saltform affect the calibration? So let's say you are producing a calibration curve using differently concentrated solutions of an acetate salt of a peptide. Later on you want to use this calibration curve to determine the content of a TFA-salt sample is this problematic?
would you be able to help me clear my doubts that what factors decide the type of calibration either linear or logit calibration? and also what determines number of standards required for an assay calibration?
Nice Thanks! It is correct to say that: there is a random error even in the ratio between the two peaks because we are not working with a perfect system, and this ratio may vary slightly from time to time, but instead of simply considering the concentration of a single substance, that is what we do when we use the method of external standard, by considering the ratio of the concentration and of signal between an analyte and a standard we increase the precision of the method and hence its accuracy??? Since that variation between the ratios its considerably lower than the variation between the single measurements??? Thanks.
Yes this is correct. Usually (but not always), the greatest source of error in the analysis happens even before we use the instrument. If you can add the internal standard early in the process, any accidental variation in your unknown will equally change the internal standard. By taking a ratio of signals, you are actually cancelling out the error. Much better than making an external calibration curve. The downside is that it is sometimes difficult to find a suitable internal standard for the experiment at hand.
Good video I have a few questions. Wouldn't Potassium work better as an Internal Standard when detecting Sodium as it is from the same column in the periodic table? Wouldn't it be also useful to run different concentrations of the sample and internal standard and do linear regression to get a better estimate for the unknown sample quantity?
The choice of internal standard depends mainly on the sample, as well as the method of detection. You first need to make sure that the sample doesn't have any of your internal standard it. As for the detector, it needs to give you a clear signal for both the unknown and the analyte. The examples I chose were somewhat arbitrary in this video. As for the regression analysis, definitely that will help. Although if you chose a standard concentration with signal similar to that of the unknown analyte, you probably won't gain much on the regression. You would still be better to prepare replicates for your analysis. And of course, there is a trade off here in that it will take more time and effort to run more samples. The calculations for regression with internal standards aren't too complicated, but they are more involved than the more basic example shown here. Thanks so much for your questions!
I have a question. What if you do a sample preparation step, this will alter the concentration right? So after you quantify using a calibration curve after the sample prep. How do you get back to the real concentration of the sample before the sample preparation step?
You should think of it like a two part problem. First focus on the concentration of the solution that you are actually measuring. After that, you need to work backwards to the original solution, usually with an equation C1V1=C2V2. You should also make note of whether the sample preparation steps will dilute the sample. Or if it will concentrate it.
Nice video...what if i have a multi-component mixture comprising 9-10 organic compounds? I guess I should prepare a calibration multi-component mixture with certain concentrations of each of them employing also a certain concentration of a good solvent, e.g. THF and of a good IS, e.g., n-octane, right? But i am not sure if i should add also water in that mixture (since I have water in my unknown samples) , would FID detect water too?
It depends how similar the analytes are, and also depends how accurate you want your measurements to be. A simple 1 point calibration, assuming every compound has equal response, will give an approximate answer. Or, as you mentioned, you can spend a lot more time to calibrate your system. As a rule, water and GC don't really mix. We make every effort to ensure no water is injected into our GC systems.
Hi! But, the use of an internal standard doesn't compensate the matrix effect, does it? For instance, in LC-MS with a matrix effect of enhancement or suppression of the ionizacion of some analytes, the construction of a calibration curve with the ratio of analyte/internal standard prevents that for quantification you have to do the standard addition method?
An internal standard will not eliminate matrix interference. However, the assumption is that the internal standard signal will be altered by that matrix in the same way as our analyte. SO the ratio of signal (internal standard vs analyte) will allow you to compensate for the matrix. In other words, you don't need a standard addition to compensate for a matrix effect. The downside of the internal standard method is that finding the perfect internal standard can be difficult. It not only needs to behave identically to your analyte, but also your detector needs to be able to measure a distinct signal for both the internal standard and the analyte
Sir for making a calibration curve using internal standard do we need to take the average of three runs. I mean do we need to take three sets of runs for each concentration??
No, you certainly do not need to have replicate runs at each concentration. Doing so would take more time to prepare, and cost more money as well. Yes, you may end up with a better result. But it depends how much effort you want to put into it. This is the same reason why calibration curves tend to use 6 or so standards. More is better, but also more effort.
Sir I have mailed my doubt to you as you have told if we have doubt we can mail so I request you if you can please clear my doubt through mail. I have sent you mail in your mail id which was provided in the video. Thanks
Sir how did you find the concentration of unknown first time as you mentiin it 0.1 M for heptanol which you said is unknown and then you put its value 0.1 M to find the response factor how wil we find first time concentration of heptanol
With an internal standard, your problem will always involve 2 different species (for example hexane and pentanol, with hexane as int std). The key is that they are chemically different and each can be measured to give their own unique signal. Ultimately, it is the "ratio" of these signals that you want. While there are two chemicals (and two signals), you will also find that each compound will be measured twice (under different conditions). In one of these instances, both concentrations (hexane and pentanol) are known. They are given to you on the problem. In the second case, only the internal standard is given. The missing concentration is what you are trying to solve. I realize this answer is a bit confusing to follow. It's hard to explain this in plain language. Thanks for your comment.
I just wanted to thank your entire family! Your brother Greg helps me train harder the last time, and you've helped me study harder than last time for my chemistry exams.
Watching this while working on A. Chem on my way to get a forensics degree! Glad you brought up how important this aspect is for the field!
Please keep going sir, your videos are very very helpful!
you're very clever, i reall do appreciate the work,thanks from italy
This video was fab really. Thank U so much
Such a great video. help a lots.
Great explanation! Thanks
Very useful and satisfying 💖
Excelente video, Thank you
Thank you for the very good explanation, I have learned a lot. Just one quastion: is the response factor (f), that is 1.4 in this example, always constant over the whole concentration range? Meaning, can we assume that is it is 1.4 at 0.1/0.1 M/M, it will also be 1.4 at 0.01/0.01 and 1/1 M/M respectively?
thank you for making this video, can you help us with calculations in IR spectroscopy?
Thank you!
Thanks a lot I I’m studying in another language and I couldn’t figure out what the hell that was about, I got it now.
very helpful .
I have a question: Does the saltform affect the calibration? So let's say you are producing a calibration curve using differently concentrated solutions of an acetate salt of a peptide. Later on you want to use this calibration curve to determine the content of a TFA-salt sample is this problematic?
would you be able to help me clear my doubts that what factors decide the type of calibration either linear or logit calibration? and also what determines number of standards required for an assay calibration?
Thank you, sir
Thank you
Nice Thanks! It is correct to say that: there is a random error even in the ratio between the two peaks because we are not working with a perfect system, and this ratio may vary slightly from time to time, but instead of simply considering the concentration of a single substance, that is what we do when we use the method of external standard, by considering the ratio of the concentration and of signal between an analyte and a standard we increase the precision of the method and hence its accuracy???
Since that variation between the ratios its considerably lower than the variation between the single measurements???
Thanks.
Yes this is correct. Usually (but not always), the greatest source of error in the analysis happens even before we use the instrument. If you can add the internal standard early in the process, any accidental variation in your unknown will equally change the internal standard. By taking a ratio of signals, you are actually cancelling out the error. Much better than making an external calibration curve. The downside is that it is sometimes difficult to find a suitable internal standard for the experiment at hand.
Good video I have a few questions. Wouldn't Potassium work better as an Internal Standard when detecting Sodium as it is from the same column in the periodic table? Wouldn't it be also useful to run different concentrations of the sample and internal standard and do linear regression to get a better estimate for the unknown sample quantity?
The choice of internal standard depends mainly on the sample, as well as the method of detection. You first need to make sure that the sample doesn't have any of your internal standard it. As for the detector, it needs to give you a clear signal for both the unknown and the analyte. The examples I chose were somewhat arbitrary in this video.
As for the regression analysis, definitely that will help. Although if you chose a standard concentration with signal similar to that of the unknown analyte, you probably won't gain much on the regression. You would still be better to prepare replicates for your analysis. And of course, there is a trade off here in that it will take more time and effort to run more samples.
The calculations for regression with internal standards aren't too complicated, but they are more involved than the more basic example shown here.
Thanks so much for your questions!
I have a question. What if you do a sample preparation step, this will alter the concentration right? So after you quantify using a calibration curve after the sample prep. How do you get back to the real concentration of the sample before the sample preparation step?
You should think of it like a two part problem. First focus on the concentration of the solution that you are actually measuring. After that, you need to work backwards to the original solution, usually with an equation C1V1=C2V2. You should also make note of whether the sample preparation steps will dilute the sample. Or if it will concentrate it.
Nice video...what if i have a multi-component mixture comprising 9-10 organic compounds? I guess I should prepare a calibration multi-component mixture with certain concentrations of each of them employing also a certain concentration of a good solvent, e.g. THF and of a good IS, e.g., n-octane, right? But i am not sure if i should add also water in that mixture (since I have water in my unknown samples) , would FID detect water too?
It depends how similar the analytes are, and also depends how accurate you want your measurements to be. A simple 1 point calibration, assuming every compound has equal response, will give an approximate answer. Or, as you mentioned, you can spend a lot more time to calibrate your system.
As a rule, water and GC don't really mix. We make every effort to ensure no water is injected into our GC systems.
i finally understand
Thank you so much sir... Very nicely explained
How does it calibrate anything?
Hi! But, the use of an internal standard doesn't compensate the matrix effect, does it? For instance, in LC-MS with a matrix effect of enhancement or suppression of the ionizacion of some analytes, the construction of a calibration curve with the ratio of analyte/internal standard prevents that for quantification you have to do the standard addition method?
An internal standard will not eliminate matrix interference. However, the assumption is that the internal standard signal will be altered by that matrix in the same way as our analyte. SO the ratio of signal (internal standard vs analyte) will allow you to compensate for the matrix. In other words, you don't need a standard addition to compensate for a matrix effect. The downside of the internal standard method is that finding the perfect internal standard can be difficult. It not only needs to behave identically to your analyte, but also your detector needs to be able to measure a distinct signal for both the internal standard and the analyte
Sir for making a calibration curve using internal standard do we need to take the average of three runs. I mean do we need to take three sets of runs for each concentration??
No, you certainly do not need to have replicate runs at each concentration. Doing so would take more time to prepare, and cost more money as well. Yes, you may end up with a better result. But it depends how much effort you want to put into it. This is the same reason why calibration curves tend to use 6 or so standards. More is better, but also more effort.
Sir I have mailed my doubt to you as you have told if we have doubt we can mail so I request you if you can please clear my doubt through mail. I have sent you mail in your mail id which was provided in the video. Thanks
Sir how did you find the concentration of unknown first time as you mentiin it 0.1 M for heptanol which you said is unknown and then you put its value 0.1 M to find the response factor how wil we find first time concentration of heptanol
With an internal standard, your problem will always involve 2 different species (for example hexane and pentanol, with hexane as int std). The key is that they are chemically different and each can be measured to give their own unique signal. Ultimately, it is the "ratio" of these signals that you want.
While there are two chemicals (and two signals), you will also find that each compound will be measured twice (under different conditions). In one of these instances, both concentrations (hexane and pentanol) are known. They are given to you on the problem. In the second case, only the internal standard is given. The missing concentration is what you are trying to solve.
I realize this answer is a bit confusing to follow. It's hard to explain this in plain language. Thanks for your comment.