Manual Baseline Correction (1D spectroscopy only)
Fitting a 7th degree polynomial by eye certainly requires a great amount of skill, yet:
- this skill can be acquired with self-teaching;
- you can stop at a much lower degree;
- iNMR contains years of experience in this field.
First we describe the ideal scenario. Later on the discussion on difficult cases. Conceptually, baseline corrections follows phase correction, but iNMR lets you free to alternate them.
- You choose the menu command ‘Process/Manual Baseline’. The spectrum is expanded and 8 vertical sliders appear. A vertical red line is drawn from the top to the center of the window: it marks the center of the spectrum.
- Use the up and down arrow keys on the keyboard to bring the baseline at the center of the screen; use the plus and minus keys to make it more intensely visible.
- Use the central slider marked with “0”. Move the thumb and an horizontal red line will move on the screen.
- When the horizontal line crosses the spectrum at the frequency of the vertical line it means that you have already found the zero-order coefficient.
- Now adjust the other central slider (marked with “1”). The horizontal line becomes a slope. Try superimposing it to the very central part of the spectrum. It's better to fit accurately a narrow region than to fit loosely a wide region.
- You can adjust other coefficients in increasing numerical order, each time correcting a larger region.
- When a thumb reaches the end of the run, or whenever you like, hit the “Subtract” button. Click it also when you have done, before closing the window, otherwise your work will be lost!
The Origin of the Cartesian System
The trick in bringing the difficulty of the task down to a human level consists in choosing the ideal central point, the one at which x = 0. Quite often this ideal point coincides with the center of the spectrum. This is also, quite often, the lowest point in the spectrum. You can specify another frequency. Put a single vertical mark, by Cmd-clicking on the spectrum, before opening the manual correction dialog. Try selecting the lowest point in your spectrum.
A Peak Falls Exactly at the Center
In this case even the first step (namely, zero-order correction) becomes difficult, because there is less visual feedback. You can either move the vertical red line, as shown above, into a unpopulated area, or visualize, interpolating into your mind, the baseline without that central peak. You fit your mental baseline at the center with the zero order correction. When you pass to the first order correction, instead of looking for the perfect fit, aim at making the moving red line parallel to the spectrum. At this point, return to the zero-order correction and aim for the exact match.
By now you have certainly realized that higher order coefficients have no effect near the vertical red line; only the zero-order coefficient has an effect at the exact frequency of the vertical red line. Even-order functions are symmetric, while odd-order functions are anti-symmetric. When you will have interiorized all of this, baseline correction will become a fun!
Learn to Give up!
It's not the case of being a perfectionist. It is enough to correct the regions around the peaks of interest, there is no further advantage in correcting the whole spectrum. In proton 1D spectroscopy you can display the integral curves: if they begin and end with two plateaus, then your baseline is very good indeed! If you are a beginner, you can limit yourself to second or third order polynomials.