Two-dimensional NMR experiments obtained on our 500 MHz NMR spectrometer have frequently been rendered useless by the presence of bands of “noise” that resemble a reflection of the desired spectrum. 2D NMR spectra often contain artifacts as a consequence of a variety of causes, such as, errors in the flip angle and phase of excitation pulses, RF inhomogeneity, imprecise or inappropriate interpulse delays, quadrature image due to a mismatch in the phase or amplitude balance of the dual detectors used for quadrature detection, general environmental instability (e.g., temperature), and artifacts arising from nonideal application of the discrete Fourier transform (1-4). The general effects of these instabilities and imperfections have commonly been discussed in terms of “ti” and “t2” correlated noise and true quadrature images resulting from incomplete suppression of N-type or P-type signals. In the case illustrated by Fig. 1, we have determined that the “noise” signals are due to interference signals originating within the spectrometer console. These types of artifacts can appear in all types of 2D ‘H NMR spectra (e.g., COSY, NOESY, TOCSY, RCT COSY) and are unrelated to the method used for quadrature detection during the incremented time domain (e.g., phase modulation (5), hypercomplex reconstruction (6), time proportional phase incrementation (7,8)). We have also observed these artifacts in experiments where no attempt was made to provide quadrature detection during the incremented time domain. Here we illustrate the source of this artifact and present a simple means by which these spurious but real signals can be eliminated. High-resolution NMR spectrometers usually employ a superheterodyne type receiver. A superheterodyne receiver converts the NMR radiofrequency signal to an intermediate frequency (IF), prior to final phase detection, by mixing a local oscillator (LO) frequency with the NMR signal to produce a difference frequency which falls within the bandwidth of the IF amplifier circuit. There are, however, two, RF frequencies which will combine with a given LO frequency to produce the same difference frequency (9, 10). For example, an NMR frequency of 500 MHz and a LO of 506 MHz will produce an IF frequency of 6 MHz. An RF of 5 12 MHz (the image frequency) when mixed with the 506 MHz LO would also produce a frequency at the 6 MHz IF frequency. In high sensitivity superheterodyne RF receivers one of the main limiting factors determining their performance is a process called reciprocal mixing. In this process the noise present on the receivers own LO is converted into the bandpass of the IF
Read full abstract