Skip To Content

The Basics of Interpreting a Proton (1H) NMR Spectrum

by Sanji Bhal, Director, Marketing Communications

1H NMR is the go-to technique to help identify or confirm the structure of organic compounds or those that contain protons. A solution-state proton spectrum is relatively fast to acquire, compared with other nuclei, and a lot of information about the structure of a compound can be deduced from it.

With centuries of combined experience in NMR data interpretation, we thought we’d share the basics to help those just starting out with analyzing solution-phase 1H NMR spectra.


The Anatomy of a 1H Spectrum

What is chemical shift?

This is the relative position of proton peaks on the horizontal axis of an NMR spectrum and one of the ways in which structural information is extracted from NMR spectra. Electron density around the proton results in shielding/deshielding effects, so different types of protons (aliphatic, aromatic or aldehydic) present in different ranges of chemical shifts.

Electronegative groups attached to -CH decrease electron density around the proton, leading to deshielding so the chemical shift increases. Tetramethylsilane (TMS) is a common standard used in NMR in which the proton is highly shielded and attributed a shift of 0 ppm. The chemical shift (or frequency of resonance) of protons for the given sample are then expressed as delta (δ) values relative to TMS (rather than absolute Hz).

Types of functional groups δ value
Amine 1–5
Alcohol 1–5.5
Carbonyl 2–2.7
Ether 3–3.4
Alkene 4.5–7.5
Aromatic 6–9
Aldehydic 9–10

What is a peak integral?

The area under the peak of a proton signal is proportional to the number of equivalent hydrogen atoms giving rise to the peak. The 1H NMR spectrum of a compound with one methyl group (CH3), one methylene (CH2), and one methine (CH) will have 3 peaks with peak ratios of 3:2:1.


What does peak splitting in a proton NMR spectrum tell you?

In NMR this is called multiplicity and it provides information about the number of proton neighbors for a particular proton atom or group of atoms (neighboring typically refers to  atoms within 3 bonds). The resonance signal or peak is split by the interaction of its magnetic field with that of neighboring non-equivalent hydrogens.

Each proton resonance (or peak) is split into N+1 where N is the number of non-equivalent neighboring protons that the atom couples to. If there are no neighboring hydrogen atoms, the resonance will be a single peak (singlet, s).

A peak that is split to any degree is called a multiplet and these are some examples:

  • If there is one neighboring hydrogen atom, the resonance is split into two peaks of equal size (doublet, d)
  • Two neighboring hydrogen atoms will split the resonance into three peaks with the ratio 1:2:1 (triplet, t)
  • The splitting for a hydrogen atom with two non-equivalent neighboring hydrogen atoms will be two doublets of equal size (double doublet, dd)

What is the roofing effect?

When there is a strong coupling effect (interaction of the magnetic fields) between hydrogen atoms, the peaks can become distorted as their chemical shifts get closer together. The parts of the peak closest to each other become larger while the outer parts become smaller. The signals appear to lean towards each other, and this is known as ‘roofing’. This effect can be helpful in data interpretation to help identify neighboring proton peaks in a spectrum.


Why are there unassigned peaks in my NMR spectrum?

There are two reasons why you may have unassigned peaks in an NMR spectrum:

  1. Another material or impurity is present in your sample; giving rise to additional peaks. Further indication of this is that the integrals of some peaks don’t fit with the integral values of the majority of peaks
  2. The sample is not the chemical structure you expected.
    • It can be challenging to interpret a spectrum that doesn’t fit the expected chemical structure and our own bias in this regard can be a hindrance.
    • If your sample was the result of a reaction, once you confirm that it contains only one compound it can be beneficial to go back to your chemistry and see what unexpected transformation, reaction, or rearrangement might have occurred.
    • Spectral prediction software can also be helpful in investigating the fit of a structure to a spectrum. Software that will provide recommendations on the structures that best fit the spectrum will eliminate your bias entirely.

There may be artifacts or errors in your spectrum due to sample problems (low concentration, undissolved particulates, etc.), inappropriately set instrument parameters, and/or inadequate processing. These are not covered in this post.

Why are there broad peaks in my NMR spectrum?

A molecule with exchangeable protons or those that form hydrogen bonds (such as -OH and -NH) will often appear as broad peaks in a proton spectrum. Not only will these peaks appear at different chemical shifts depending on the solvent you use, but they can also obscure peaks from your sample. A drop of D2O will remove these signals and allow you to interpret a spectrum more easily.

For a 1H-NMR spectrum to be consistent with a chemical structure, the following must all be true:

  • All chemical shifts are in the appropriate range for the assigned protons’ local magnetic environment
  • All peaks exhibit the appropriate multiplicity
  • All peaks exhibit an appropriate integral value
  • There are no unassigned peaks (excluding water, solvent, or chemical shift standard)

If you’re interested to learn about our software for NMR data processing and analysis, read about Spectrus JS, NMR Workbook Suite, and NMR Predictors.