I'm working with catalytic conversion of a substrate to a product. To see how much product I'm obtaining through NMR as a crude analysis, I made a calibration curve using mixtures of pure reactants and products purchased from chemical suppliers.
When I plot the various mixtures of reactants and products, instead of having a direct linear relationship to one another I am getting a power law fit. So when there are small amounts of product and I normalize the product peak to how many hydrogens are in the product peak, the integral of the reactant peak is very large, and vice versa. I'm wondering what is the reason that I am obtaining a power law fit for the calibration curve of two substances vs a linear fit which I thought you would expect? Thank you!
What is the value of the exponent that you got from the power-law fit? If this exponent is close to 1.0, then you are close to the linear model. I also assume that all the variables in your power-law fit are dimensionless.
In general, the power-law that you got from the experimental data is empirically-derived and, hence, should be applicable only within the experimental conditions that you used. That is to say, if you try to extrapolate the power-law predictions outside the experimental conditins, you may get unreliable results.
Hope this helps answer your question.
Professor Yehia Khalil
Yale University
USA
Thank you both for taking the time to reply!
@Dr. Päch - Yes, I am using the absolute peak areas. I don't think I can go into specifics of what substances, but yes they are 'innocent' as far as establishing equilibrium.
@Prof. Khalil- Thank you so much for this, I believe this is what I needed to hear. The value I'm getting for the fit is very close to 1, it is 1.088 actually. I noticed it was linear after plotting log x vs log y. You are also correct that the variables in the fit are dimensionless. I believe this solves it!
- Badges
- Science topic
- Similar topics
- Biological Science
- Molecular Biology
- Molecular Methods
- Cloning