The impact of poor tuning and matching

Tuning and matching is the process of optimising the frequency and resistance of the circuit that includes the detection coil. Every sample has a slightly different ionic content and so the probe should be tuned and matched for each sample. Modern NMR probes have automatic tuning and matching devices, but the Skaggs NMR Facility probes do not. Here, I look at the impact of poor tuning and matching on a 1D ¹H spectrum.

The NMR spectra were collected using a sample of cholesteryl acetate in chloroform-d using a room temperature 5 mm broad band inverse detection probe. The spectra were collected using a single scan, a one minute relaxation delay, and the zg pulse sequence that implements a single 90^o pulse followed by data acquisition. Initially, the probe was tuned for a methanol-d₄ sample, a 1D ¹H spectrum was collected, and the 90^o pulse with this tuning was measured. The probe was then tuned and matched for chloroform-d, another 1D ¹H spectrum collected with identical parameters, and the 90^o pulse redetermined.

A stackplot of the two spectra plotted with the same scaling is shown below. The lower spectrum in blue was collected with poor tuning and matching, while the upper red spectrum was collected with good tuning and matching.

Clearly, improving the tuning and matching has improved the sensitivity. Measuring the signal-to-noise on the peak at 5.40 ppm gives 777.13 for the poorly tuned spectrum and 1607.32 for the well tuned one.

The poorly tuned spectrum is less sensitive because the 90^o pulse is not calibrated correctly. Poor tuning and matching results in a longer pulse being required to rotate the magnetisation 90^o from equilibrium along the z-axis, to the x-y plane where the signal can be detected. For the well tuned spectrum the 90^o pulse length was 7.70 µs. For the poorly tuned spectrum the 90^o pulse was 24.25 µs. Both spectra were recorded with a 7.45 µs pulse. Thus, the poorly-tuned spectrum used a 28^o pulse which would give 46.4% of the signal, while the well tuned spectrum was recorded with an 87^o pulse which would give 99.9% of the maximum signal. The signal-to-noise measurements gave values with a very similar ratio.

tuning	poor	good
90o pulse	24.25 µs	7.70 µs
effective pulse	28o	87o
signal available	46.4%	99.9%
signal-to-noise	777.13	1607.32

The 5 mm probes are much more sensitive to tuning and matching than the 1.7 mm probe because the volume of solvent is so much larger. If you want to maximise your sensitivity make sure the probe is tuned and matched for your solvent.