Tutorial: How to Fit An Abstract Spin System
Don't repeat the common mistake of using a program before learning it. You risk to waste time and stress yourself.
By following this simple tutorial you can understand, in a couple of minutes only, how fitting works. The trick is to start with a perfect and trivial example, that requires the least amount of computation. In this way you can learn the iNMR commands before facing the problems connected with a real life example. First learn to walk, then learn to run. In our case, we'll fit an ideal spectrum, with no noise, no baseline and no phase distiortions. The initial part of the tutorial describes how this ideal spectrum is created.
Open iNMR (or iNMR reader). Dismiss the Open dialog. The command “Simulate/New Spectrum” creates the theoric spectrum of a simple AB system. Let's change some parameters. Set the chemical shifts to 2.5 and 0.5, the J to 20 and the population to 1e7.
Close the sheet dialog. Double click into the main window to scale-down the peaks. Change the color to black (command: “Format/Levels & Colors”). Edit the spectrometer frequency (into the drawer; it's the first item in the list). Set it to 60 MHz.
Our fake “experimental” spectrum is ready. Now create another untitled document to fit it. The command is the same (shortcut: Cmd-N). Close the sheet dialog that appears after the command. Open the color dialog instead and change the color to red. The meaning of the colors is:
- black = experimental spectrum.
- red = simulation.
We want to work with a single window. While the red spectrum is on the foreground, open the Overlay Manager (command: “Format/Overlay”). When the dialog appears, check the only visible item into its list.
Dismiss the dialog. In this moment the red spectrum sees the black one and believes it's an experimental spectrum to be fitted. To permit the fitting process, the two spectra must have the same spectrometer frequency. Do not edit it. Simply click the button “fix” (up to version 4 the name was “assimilate”)
It copies the spectrometer frequency (and a few other things) from the experimental spectrum. Press also the button on the left (refresh), to be sure that the whole simulated spectrum is on display. The intensities of the two spectra differ, however, by 4 orders of magnitude. To compensate this fact, check the parameter “pop” (fourth from the bottom), and click again the Fix button. When the population is checked, and you click on “Fix”, iNMR changes the value until the integral of the simulated spectrum equals the integral of the experimental spectrum. The integral is calculated from the left to the right ends of the visible portion of the frequency scale. It is therefore important that the scale goes from 0 to 3 ppm (to include all the experimental peaks). Double click to optimize the vertical scale. Now you should see all the peaks (red and black ones).
The chemical shifts and the coupling constant are different. Check the corresponding values into sidebar (uncheck “pop”).
The command “Simulate/Fit to Overlay” will change the selected parameters until the difference between the two spectra is minimized. That's all.
The Faster Way
We have divided the fitting process into 3 stages (equalizing the spectrometer frequency; equalizing the total area; fitting the individual lines). In this way can easily follow what's happening and verify that iNMR performs its tasks correctly. You can however simplify the work-flow, if there is less difference between the two spectra. For example, create a third untitled document, with chemical shifts 2.7 and 0.3 and span = 1000 (click the refresh button after setting the span). Overlay the “experimental” spectrum, just as you did before. After you close the Overlay Manager, check the 4 parameters at the bottom of the list, starting from the population, and give the “Fit to Overlay” command. The three stages outlined above will be performed with a single command. The fitting will not be very good. Repeat it!
This example, with all its simplicity, has shown many important concepts. Another thing you need to know is how to enlarge the spectral width of the simulated spectrum. It is proportional to the parameter called “span”.
To verify it, double the value of “span” and click “refresh”. Got it?
Now you can try by yourself to generate larger systems and fit them. When everything is clear you can begin to fit an authentic experimental spectrum of yours. You are also invited to read the next tutorial in which we fit an experimental spectrum.