HPLC quantitative analysis

[abdi22]

Hi

How do you calculate the amount of a compound from a sample ? To be more precise i do know the procedure somewhat but the problem is i've got different retention time for both of the compounds i am trying to determine i.e retention time for both compounds in the sample "X" is different than the retention time of the pure compounds. Does this mean that the compounds aren't present in the sample (sample x) ? For more details i've included the entire question below.

Sample X (serum) has gone through the following process before being injected into the HPLC instrument: 500 μl of sample X was added to 500 μl internal standard (vitamin A acetate and vitamin E acetate), and mixed well. The mixture was extracted with 1000 μl hexane and was transferred to a clean tube and then evaporated under argon. The evaporated extract was dissolved in 500 μl methanol and injected into an HPLC with UV detector. All analysis were conducted using the same HPLC conditions. With the help of the attached chromatograms determine the amount of vitamin A (mg/l) and vitamin E (mg/l) i ”Sample X” by using peak height
(For all retention times in this experiment, we allow a time window of + 0.08 minutes.)

1. Vit A chromatogram
http://tinypic.com/view.php?pic=n4z4ug&s=6

2.Internal standard(vit a & e) chromatogram
http://tinypic.com/view.php?pic=6p8dq8&s=6

3. Vit E chromatogram
http://tinypic.com/view.php?pic=2vkbwcw&s=6

4. Sample x chromatogram
http://tinypic.com/view.php?pic=2r3ikn6&s=6

I've tried to embed the chromatograms as suggested here viewtopic.php?f=5&t=12662 but i've been unsuccessful so far and that's why i've copied the URLs.

Any help would be greatly appreciated

[Don_Hilton]

From what I can see of the chromatograms, I think the retention time differences may be small - but I can hardly read the scale to get an idea of peak width. So, rather than try to figure out your chromatograms or give a bad guess:

The retention time tells you about a chemical property of the individual compounds. If you assume the chromatographic conditions are constant, the retention time will not change. (So good so far...) But!

Compounds from the sample matrix can participate in the partitioning process - making a very small change in the "local" chromatographic conditions - or in real samples, peaks may move a bit - just due to matrix on a column. (By local, I am describing that the gradient and stationary phase are nominally the same, but the concentrations of these other compounds are high enough they now figure in the solvent properties and affect the partition coefficients in the portion of the chromatogram where they occur.)

And if there is enough of the compound itself to affect the "local" chromatographic conditions, the apex of the peak will move and peak shape will be affected (overloading of the column.)

So what to do?

If you suspect that peaks in one run are the same compounds that you have seen in another run. Mix the two samples and see if the peaks overlay each other or separate in the third run. If the peaks separate in the same chromatographic run - you have a definitive answer. Peak shape is important here - if the peak in the mixed sample has a single apex, but is more broad than what was seen in the other runs, you have a clue that you may have two compunds that elute very closly - but not quite.

If you have peaks that elute at the same time, but need to ensure they are the same compound, not an accidental coelution it becomes more difficult - you can only reduce the probability that you have a wrong answer. But you can reduce that probability significantly. For a given sample you can change chromatographic conditions or detection conditions. Changes in factors that will change the partition coefficient do not change the partition coefficient at the same rate (or even in the same way) for all molecules. Thus, changes in gradient, solvents, stationary phase, etc. can help to confirm that the supposed analyte in the sample and the standard compound are likely to be identical. And with a UV detector it is easy to check to see if the known and unknown peaks have the same spectral characteristics.

The tests to try to rule out a coelution of some other compound seem to be open ended because you only have a proof when you prove the compounds to be different. So when do you stop? This depends entirely on what you are doing (and how close you are to retirement ). If you know the samples, this is easier. If you are analyzing a mixture and have confirmed the identities of the ingredients, the confirmation of the resulting mix is fairly easy. Or if you are analyzing tobacco and find a compound at the retention time of nicotine with matching spectral characteristics - I'll bet you a nice lunch on that one that it is nicotine. If you are looking for a metabolite in urine at trace levels, you are going to have to do some careful work to ensure that you are not confusing the target analyte with something else.

I hope that gives you some help - or at least some ideas to get you going.

[abdi22]

Thanks for the reply Don but what i was really after is how to calculate the amount of vitamin A & E in the sample as this was the main goal of the analysis but i really appreciate your suggestion as i will use it in my next lab experiment. The question about the retention time being different was more of a side question to myself so that i could reflect upon it as what i was taught in the lectures was that it was supposed to be the same, so at least i now have a better understanding of it and also a solution that i can use to confirm if they are the same or not.

So my next question is how do you move on from there when trying to to solve for the concentration of both vitamins? I know that i've to set up a calibration curve but this is where i got stuck and i need help.

Vit E (30mg/l)
RetTime Width Height
3.836 0.0867 30.26071

Vit A (1mg/l)
RetTime Width Height
2.609 0.0661 29.90904

Internal Standard (mixture of Vit A & E) 500µl
RetTime Width Height
1. 2.867 0.0833 10.43465 ---> for vit A
2. 4.369 0.1001 3.08223 ---> for vit E

Sample X
RetTime Width Height
1. 2.288 0.1015 1.80386
2. 2.608 0.0657 9.71365
3. 2.858 0.0947 6.55282
4. 3.783 0.0853 6.10469
5. 4.295 0.1108 1.37299


So the next step would be drawing a calibration curve for each of the vitamins ? so i should plot peak height(y-axis) against concentration(x-axis) but my question is how do i find the concentration in both the sample and internal standard? The only concentration given is only for the pure compounds ie 30 mg/l for vit E & 1.0 mg/l for Vit A. I've tried using the formula C1*V1= C2*V2 to find the concentration of sample and IS but since we haven't be given the volume of any of the pure compounds (vit a & e) , i can't use this formula since we are dealing with 2 uknowns instead of only 1 uknown. Is there another way of determining the concentration of vitamins in the sample ??

For vit E we know the following
1. concentration before it is added to any sample or mixture which is 30mg/l
2. Retime of both when pure and in mixture, but when it comes to the sample X which retTime should i use? should i just assume that the last two retTimes( peak # 4 & 5) belong to Vit E.

For vit A
The same problem applies here too since we've got different retention time in all cases i.e in pure compound, internal standard & sample X.

So with that in mind how should i proceed forward in finding the concentration of these 2 vitamins in sample X?

[Don_Hilton]

I am at a bit of a loss at what you have for the various injections. When using an internal standard, one includes internal standard and analytes in the same run and obtains an area ratio. Doing this for multiple levels of analytes, one can plot a response ratio as a function of concentration. One also has the internal standard and analytes in the sample and obtains an area ratio. That ratio may be used with the plot generated above to determine the conentration of the analytes.

You need some information about the addition of internal standard to the sample. In many cases samples and standards are made up with a fixed quantity of standards and samples and both samples and standards are diluted to the same volume.

Is this an experiment for which you have instructions?

[abdi22]

I am at a bit of a loss at what you have for the various injections. When using an internal standard, one includes internal standard and analytes in the same run and obtains an area ratio. Doing this for multiple levels of analytes, one can plot a response ratio as a function of concentration. One also has the internal standard and analytes in the sample and obtains an area ratio. That ratio may be used with the plot generated above to determine the conentration of the analytes.

You need some information about the addition of internal standard to the sample. In many cases samples and standards are made up with a fixed quantity of standards and samples and both samples and standards are diluted to the same volume.

Is this an experiment for which you have instructions?

No it is a question that i came across from my lab manual, so i was just trying to see if i could solve it. The main task of the question is to find the amount of concentration of the vitamins in the sample by using the peak height. Other than details in the chromatogram no further information is given apart from details in the question itself about how the preparation was made. Anyway thank you for you input but one last question. Do you happen to know of websites/online books that explain in detail about determining the concentration of a compound ? Ive gone through the following website http://www.lcresources.com/resources/getstart/index.htm but it doesn't explain in detail

[Don_Hilton]

I think I encountered making calibration curves with internal standards way back when I took a class in instrumental methods of analysis. As I recall we did very basic calibration plots (no computers - except the mainframe, which was used only for computer classes and fancy research projects). I've picked up the rest of what I know in the lab or from interesting papers over the years...

If you are at a college or university, see what is being used in the instrumental analysis class.

Or, someone who has used these materials on line may be able to jump into the discussion and help out.

[abdi22]

I think I encountered making calibration curves with internal standards way back when I took a class in instrumental methods of analysis. As I recall we did very basic calibration plots (no computers - except the mainframe, which was used only for computer classes and fancy research projects). I've picked up the rest of what I know in the lab or from interesting papers over the years...

If you are at a college or university, see what is being used in the instrumental analysis class.

Or, someone who has used these materials on line may be able to jump into the discussion and help out.

Thanks and the first thing i tried was my class notes, lectures etc but the only thing covered was only the separation principles of different instrumental analysis(basic stuff really) and it was only a 2 hr lecture. So everything else you had to make your own notes. Nothing was mentioned about how you were supposed to find concentration of a compound using a chromatogram & the only time i came across it was during lab practicals. We worked in groups, so different groups got different experiments to run and i was unfortunate enough to get a different experiment to work with. I also talked to one of my lab assistant's and she told me that questions dealing with concentration calculation is a typical exam question. I've also tried to get in touch with the groups who did this particular lab practical but have been unsuccessful and that's why i resorted to asking online and by sheer luck i came across this forum. Anyway i am really thankful for your help

[tom jupille]

This may help:
http://www.lcresources.com/resources/getstart/4e01.htm

[MaryCarson]

abdi22:

You don't have enough standards to do a calibration curve. You are doing a single-point calibration calibration. There is an implicit assumption that the calibration is linear through the origin. The calculation is all just ratios. Try rearranging your data into this table and see if it makes more sense to you. You only need the height info for the calculation--the retention times are just there to make sure you got the right peak ID.

injection | known conc. | Vit A RT | Vit A ht | Vit A IS RT | Vit A IS ht | Vit E RT | Vit E ht | Vit E IS RT | Vit E IS ht |

Vit A | 1.0 | | | | | | | | |

IS | ---- | | | | | | | | |

Vit E | 3.0 | | | | | | | | |

Sample | TBD | | | | | | | | |

Hint: you don't need to know the concentrations of the acetates in the IS as long as it went through the same procedure.

[MaryCarson]

Well, so much for my carefully spaced out vertical bars to represent a table...

[abdi22]

abdi22:

You don't have enough standards to do a calibration curve. You are doing a single-point calibration calibration. There is an implicit assumption that the calibration is linear through the origin. The calculation is all just ratios. Try rearranging your data into this table and see if it makes more sense to you. You only need the height info for the calculation--the retention times are just there to make sure you got the right peak ID.

injection | known conc. | Vit A RT | Vit A ht | Vit A IS RT | Vit A IS ht | Vit E RT | Vit E ht | Vit E IS RT | Vit E IS ht |

Vit A | 1.0 | | | | | | | | |

IS | ---- | | | | | | | | |

Vit E | 3.0 | | | | | | | | |

Sample | TBD | | | | | | | | |

Hint: you don't need to know the concentrations of the acetates in the IS as long as it went through the same procedure.

Hi

Thanks for the reply but i did figure it out eventually that it was a single point calibration & determining the amount of analyte present in the sample was fairly easy. At first what i didn't understand was why the analyte had different Ret-times as this somewhat contradicted what i was told & also what i read from some books. I thought that the RetTime had to be the same inorder for us to able to identify it so when i looked at the chromatograms and saw a difference in the Ret-Time, i was confused at first. I didn't know how i could be able to verify that it was the same analyte and not something else.

Since the course i was taking was meant to be an introductory one, i figured that for now i should just assume that it is the same analyte due to fact that the difference was only minimal but that might not the case every time. At least for now i do have a better grasp of it and i would like to thank all those who have contributed to this thread.

Thank you!!