Multiple Component Calibration Curve in Empower.

[DylonHussain]

I need a way to put 6 different components on a calibration curve in Empower.

I was given a method to test the linearity of the FID on an agilent GC that consists of one injection of a sample with 6 different saturated hydrocarbons. The method has a spec on R^2 value of the response factor (I'm fairly confident that it means to find the R^2 for concentration vs area). Finding the R^2 value for this requires a best fit curve, which involves least squares minimization, so I believe that performing this calculation in a custom field is out of the question. Empower cal curves can display R^2 values so I think that this would be the easiest option. So does anyone know of a way to make empower ignore that the peaks are different components and treat it as a multilevel calibration of one component?

Thanks in advance,

Dylon

[shaun78]

Use the component editor in the sample set to add all the components and their respective concentrations.

In the processing method name all components with their corresponding retention time. Create a named group that contains all of your components. Have Empower plot the calibration curve for the named group.

The above should set you up. Let me know if you need any more information with that and if I can be of further assistance.

[DylonHussain]

I created the named group with all of the components and processed but it did not generate a calibration curve for the group. I clicked the "calibration curve" button in the view result window and all of the components had a curve but the group I made did not. It says "Invalid Curve. Do not use for quantitation" in that window. For the named group I have "Source of Calibration X Value" set to "User Entered, Curve or Sum Peaks for Quantitation" and "Reported RT" set to "None." Is this correct?

[shaun78]

Dylon,

I apologize…

It sounds like you have everything setup correctly.

In the component editor of the sample set you only need to enter the Named Group as the compound and the corresponding concentrations.

I believe the above should resolve your problem. Please let me know if it does not.

[DylonHussain]

So the name of the group that I made is "Linearity." I tried adding linearity after the six components without a value and it disappears when I try to save it. I tried adding linearity after the six components with a value of 1 (arbitrary choice that was not the same value of any of my components) and it gave me a calibration curve where the Y value is the sum of my components and the X value is one. I tried adding six components (all called Linearity) with the concentrations in my standard but it only saves the first one.

I my not have been clear in my explanation of what I need to do: I need Empower to make a multi-level calibration curve based on a single injection with six different components, treating each of the different components concentrations as separate x values on the same cal curve.

Thanks again for your help!

[DylonHussain]

I was able to successfully calculate R^2 as the covariance squared divided by the product of the variances in a custom field.

[Synne Bulinckx]

Hey, I'm struggling with a similar idea, would you be able to share how your custom field looked? Did you have to make custom fields for each parameter (covariance, standard deviation, etc.) separately?