1.11. How To Predict NMR Spectra

It is possible to analyze a chemical structure and predict the ¹H NMR and/or ¹³C NMR spectrum. This combines all previous skills but is normally straightforward with practice. There are three common ways of asking students to perform this task: predict signals and choose from possible spectra, predict signals and report values, or predict signals and draw the spectrum.

Identify in some way which sets of hydrogen atoms are chemically equivalent. See Section 1.6.3. The speed of this step often improves dramatically with practice and may eventually be done internally (in your mind’s eye). Remember to pay attention for diastereotopic hydrogens (not present in this example). This text uses colour-coding but a common alternative would be to use number- or letter-coding for each group of hydrogens.

For each set of chemically equivalent hydrogens predict ROUGHLY what the chemical shift would be (see Figure 1.24). You do not need to be exact. If nearby electronegative atom(s) might raise the chemical shift outside of the normal range make a note of it.

For each set of chemically equivalent hydrogens predict multiplicity (see Section 1.9.4). Remember to pay attention for diastereotopic hydrogens (not present in this example).

For each set of chemically equivalent hydrogens predict the integration value. You do not need to reduce any values, just use the number of chemically equivalent hydrogens in each signal.

If you are comparing your prediction with a set of possible spectra, simply find the one that best matches your prediction. Remember that your chemical shifts are approximate. Remember that multi-coupled signals have variability in their shapes. Remember that signals with similar chemical shifts can overlap.

A common approach is to systematically check for each of the signals predicted:

Could there be a signal around the 2.0-2.6 chemical shift range, with an integration of 3, that is a singlet?

If no, probably wrong spectrum. If yes, then…

Could there be a signal around the 4.5-6.5 chemical shift range, with an integration of 1, that is a doublet?

If no, probably wrong spectrum. If yes, then…

etc.

If you are reporting the data, list the signals in order from smallest to largest chemical shift with the multiplicity and integration for each signal in a bracket immediately after the chemical shift. In laboratory or professional settings ranges are not reported, only specific chemical shifts. However, since the exact shift is not known this is acceptable.

example:

δ 0.9-1.2 (s, 18H), 1.4-1.9 (t, 1H), 1.5-2.5 (dd, 2H), 2.0-2.6 (s, 3H), 4.5-6.5 (d, 1H), 4.5-6.5 (dt, 1H)

If you are drawing the spectrum add signals one-by-one to the image. When drawing spectra remember that this is chemistry, not art; the drawing does not have to be beautiful or flawless, only legible and accurate as reasonably possible.

Approximate the chemical shift as best you can, typically by choosing the middle of the range for the shift. If multiple signals would have similar chemical shifts choose one to be slightly higher (but still within the range) and draw them so they do not overlap. If they are very close together, consider indicating in some way which lines come from which signals. At an introductory level it is not required to predict which of the signals would have the higher/lower chemical shift value.

Draw the signal as the predicted multiplet with correct intensity ratios as best as possible. For example, a triplet should have 1:2:1 intensity ratio. The left and middle drawings are obviously not 1:2:1. The right drawing is not perfect but acceptable.

For signals that are complex (e.g. a dt) draw something that is easily identifiable as the correct multiplet (e.g. two triplets very close to each other). Do not worry if the “real” shape would be more complicated.

Draw the signal as the predicted multiplet with approximately correct integrations (relative sizes) as best as possible. Do not worry about being perfect. It is often helpful to write the integration for a signal above the drawing of it to make it clear for readers.

Example:

This task is about applying and understanding the concepts more than perfect accuracy. For example, drawing the green signal at a higher shift than the red one would also be acceptable (and more accurate) but both would be graded as correct. In the actual ¹H NMR spectrum for this compound the green and orange signals have more complicated shapes. Again, this is not important in the drawing. The orange signal has an integration of 2, but is not perfectly twice the size of the pink signal (integrating to 1) nor 2/3 the size of the blue signal (integrating to 3) beside it. However, it is visibly larger than the pink and smaller than the blue, so this is acceptable. Do not worry about creating a perfect picture, focus on conveying understanding and the information you predicted.

The exact same approach would be used to predict ¹C NMR spectra. These are typically faster (no multiplicity, integration (relative line size) is not meaningful).