Wednesday, January 11, 2017

T1 and T2 relaxation

In NMR, relaxation is the process by which magnetization returns to equilibrium. Relaxation is important because it determines how long to wait between scans, how quickly signals decay, and the width of peaks in a spectrum. In the vector model relaxation is separated into T1 relaxation along the z-axis and T2 relaxation in the x-y plane.

The vector model is a visual depiction of the net magnetization produced by the 1020 or so identical nuclei in an NMR sample. It represents the average of the magnetization produced by an ensemble of nuclei that give rise to a single peak in a spectrum. At equilibrium in a magnetic field the net magnetization of the NMR-active nuclei will be aligned with the +z-axis. After a 90o pulse the net magnetization vector will lie in the xy-plane and will start to relax back towards the z-axis.

The figure below shows the net magnetization vector after a 90o pulse and resolves it into two vectors; one in the xy-plane and another along the z-axis. As the magnetization relaxes towards equilibrium the vector in the xy-plane becomes smaller as the vector along the z-axis increases. Eventually, the vector in the xy-plane becomes zero and the vector on the z-axis reaches its equilibrium magnitude. The increase in net z-magnetization is known as longitudinal relaxation and occurs with time constant T1. This process is mainly caused by the interaction of the spins with their surroundings and so is also known as spin-lattice relaxation.

The T1 time constant is what determines how long to wait in between scans when recording an NMR spectrum. To obtain maximum signal one needs to wait until the magnetization has returned to equilibrium along the z-axis. Relaxation closely approximates an exponential process so after waiting one time constant, 63.2% (i.e. 1 - e-1) of the magnetization is recovered. T1s are typically in the range of 1-10 s, so a relaxation delay of two seconds is a reasonable value.

After a 90o pulse the net magnetization vector is left rotating in the xy-plane. Since this vector is the sum of the magnetization of many different spins, if these spins experience slightly different magnetic environments (e.g. shimming or temperature variations over the sample volume) their rotation frequencies will be slightly different. This will cause the individual components of the net magnetization to spread out and a reduction in the size of the net magnetization in the xy-plane. This process is known as transverse relaxation and occurs with time constant T2. Since it is mainly caused by the interaction of the spins it is also known as spin-spin relaxation.

The T2 time constant determines how quickly the FID decays and thus the width of the peaks in an NMR spectrum.  Faster T2s produce broader peaks. T2s can range from 10s to 10ms.  Large molecules with short T2s produce FIDs that decay rapidly and so acquisition times should be limited to prevent adding noise to the acquired signal.

The rate of T1 and T2 relaxation depends on the magnetic field, ω, and the correlation time of the molecule, τc. T2 decreases as ωτc increases, but T1 passes through a minimum around ωτc = 1. The T1 minimum is what separates "small molecule" NMR from biomolecular NMR. Most small molecule samples are on the left of the T1 minimum where T1≈ T2 and one does not have to worry about signals decaying too quickly. For large biomolecules where T1 > T2, transverse relaxation becomes the dominant factor in designing experiments.


  1. Again, fine succinct summary of a fundamental physical constants in NMR. Interesting how one develops a 'gut' feeling for T1 and T2 after seeing enough FIDs on the computer screen! Perhaps, in a future blog, you could share other factors that influence relaxation such as paramagnetic species (triplet O2, transition metals with unpaired electron spins...etc. No Chromerge® for cleaning NMR tubes!)

    1. Thanks Ted. All suggestions for future blog posts are welcome! I will look into those perhaps undesired methods of relaxation enhancement.