Chemical Exchange on the NMR Timescale

Introduction

Molecules can interconvert between environments while an NMR instrument observes them. A proton may move between conformers, a ligand may bind and unbind, or two sites may exchange through a chemical reaction. The spectrum depends on how fast that interconversion is compared with the frequency difference between the environments. Separate peaks, broad merging signals and one averaged peak are different views of the same dynamic process at different rates.

Core explanation

Imagine two sites A and B whose non-exchanging frequencies differ by Δν hertz. Exchange rates are compared with the angular frequency separation 2πΔν, with precise boundaries depending on populations and the kinetic model. In the slow-exchange regime, a nucleus spends long enough in each environment to precess at its characteristic frequency, so two peaks appear near their individual shifts. Their areas can approximate populations if pulse and relaxation conditions allow quantitative integration. Exchange may still broaden these peaks compared with a non-exchanging reference.

As exchange becomes comparable with the frequency separation, line shapes broaden and distort. This is the intermediate regime. For equally populated sites with a simple symmetric two-site model, the two lines can merge at coalescence. The coalescence temperature is often used to estimate an exchange rate, but a formula derived for equal populations and simple kinetics should not be applied blindly to unequal or overlapping signals. Fitting full line shapes over several temperatures is usually more informative.

In the fast-exchange regime, each nucleus samples both environments many times during the time needed to distinguish their separate precession frequencies. The spectrum shows one averaged resonance, approximately δ obs=p A δ A+p B δ B when the simple model applies. The peak may narrow again as exchange becomes very fast, although exchange can still contribute a finite linewidth. A single averaged peak therefore does not prove the molecule has only one static environment.

The exchange regime depends on the spectrometer field. A chemical-shift difference specified in ppm corresponds to a larger separation in hertz on a higher-field instrument. The same physical exchange rate can therefore appear slower relative to Δν at higher field, potentially revealing signals that were merged at lower field. Temperature can also alter both rate and equilibrium populations, so changes in peak position are not always purely kinetic.

Examples include hindered amide rotation, conformational ring flips, proton transfer, ligand association and chemical reactions. A site-exchange model should identify which nuclei move between which distinct environments. Merely seeing a broad peak is not enough: short T2, magnetic inhomogeneity, unresolved coupling, paramagnetism and sample heterogeneity can also broaden signals. Dynamic NMR uses controlled temperature or field variation, line-shape fitting and independent chemical evidence to distinguish exchange from these alternatives.

Fast proton exchange with solvent may suppress or broaden O–H and N–H couplings and make their integrations unreliable. A D₂O shake can show disappearance of exchangeable proton signals, but that experiment changes the chemical isotopic composition and is not a general kinetic measurement. Deuterated solvent peaks and water traces need separate assignment before attributing a broad feature to molecular exchange.

In a two-state interconversion at equilibrium, forward and reverse rates need not be equal if populations differ. Detailed balance requires p A k AB=p B k BA in the simplest closed reversible model. A fitted total exchange rate must therefore be interpreted in the context of populations and the chosen convention. Avoid quoting an activation barrier from one line width unless the model and calibration support it.

Step-by-step reasoning

Identify the two candidate environments and their slow-limit frequency difference in hertz. Compare exchange rate qualitatively with that separation. Look for separate peaks, broad coalescence or an averaged signal, then check field and temperature dependence. Exclude other broadening causes before fitting a kinetic model.

Visual explanation

Sketch a sequence of spectra as temperature rises: two narrow peaks, two broadened peaks moving into one wide feature, and one narrower peak between the original positions. The last peak sits at a population-weighted average shift, not necessarily the exact midpoint if populations differ.

Real-world analogy

Two people speaking one at a time are easy to distinguish if each speaks for a long interval. If they switch turns rapidly compared with a listener's sampling window, their voices can seem blended. This captures time averaging, though NMR line shapes follow coherent spin dynamics and specific rate equations.

Real-world example

Restricted rotation about an amide bond can make groups on either side appear inequivalent at low temperature. Warming may speed rotation, broaden the separate NMR signals and eventually produce a single averaged resonance. Temperature-dependent line shapes can then constrain the rotational barrier when other possible changes are controlled.

Why?

Why does higher field sometimes resolve exchanging sites better? Their ppm separation stays roughly the same, but frequency separation in hertz grows with operating frequency. The exchange rate in s⁻¹ need not grow merely because the magnet is stronger, so exchange becomes slower relative to the increased spectral separation.

Common misconception

Coalescence is not chemical disappearance of one site. It is a line-shape effect from interconversion during observation. Conversely, two separate peaks do not imply no exchange; the exchange may be slow on the NMR timescale yet chemically important over seconds or minutes.

Worked example

Two equally populated proton environments would appear at 2.0 and 4.0 ppm without exchange. If exchange becomes very fast and the simple population-average model applies, their single observed peak approaches (0.5)(2.0)+(0.5)(4.0)=3.0 ppm. If populations instead become 0.75 and 0.25, the average is 2.5 ppm. Thus an averaged signal's position reflects both intrinsic shifts and populations, not merely the exchange rate.

Quick check

1. Can one sharp NMR peak be consistent with two chemically distinct environments exchanging rapidly? Answer: Yes. Fast exchange can give a single population-weighted averaged resonance even when two distinct environments exist instantaneously.

Exam focus

Compare exchange rate with peak separation in hertz, not ppm alone. State the kinetic and population assumptions behind coalescence calculations. Consider relaxation and inhomogeneity as alternative broadening mechanisms, and do not infer an activation barrier without temperature-dependent data and an appropriate model.

Advanced insight

The Bloch–McConnell equations describe coupled magnetisation evolution during chemical exchange and predict line shapes across all regimes. Relaxation dispersion experiments detect some exchange processes even when individual peaks are not separately resolved. More than two states, unequal populations and coupled spins can yield behaviour far richer than the textbook pair of coalescing singlets.

Summary

Slow exchange yields separate site resonances, intermediate exchange often broadens and merges them, and fast exchange yields a population-weighted average. The classification depends on rate relative to frequency separation in hertz, which changes with field. Temperature-series line shapes and controls are needed to distinguish exchange from other broadening and extract kinetics.

Practice questions

1. What spectrum is expected for two sites exchanging much more slowly than their frequency separation? Answer: Two distinct resonances near their individual chemical shifts, possibly with some exchange broadening. 2. Two sites at 1 and 5 ppm have populations 0.25 and 0.75. What is their simple fast-exchange average? Answer: (0.25)(1)+(0.75)(5)=4 ppm. 3. Why can the same exchange process look different at higher magnetic field? Answer: A fixed ppm separation becomes a larger hertz separation, changing the ratio of exchange rate to spectral separation. 4. Does one broad line prove intermediate chemical exchange? Answer: No. Short T2, field inhomogeneity, unresolved coupling or sample heterogeneity can also broaden it.