based on the simultaneous selection in a single scan of both Recently, many new high-resolution NMR experiments orthogonal magnetization components present during an using pulsed field gradients (PFG) for coherence selection evolution period, and when applied to 1D experiments, no have been proposed which offer a much better spectral qualextra requirements in the acquisition or processing steps are ity than analogous phase-cycled experiments in generally needed. As a result, and in accordance with the theoretical shorter total experiment time (1). Specifically, the implepredictions (9-11), a maximum increase of the signal-tomentation of PFG in selective homo-and heteronuclear 1D noise ratio by a factor of 2 and 1.41 can be achieved for IS experiments (2) affords a powerful tool for obtaining highand I 2 S systems, respectively. To the best of our knowledge, quality 1D spectra without the presence of undesirable artithis is the first application of the PEP methodology in a onefacts arising from the imperfect subtraction inherent in the dimensional NMR experiment. conventional phase-cycling selection. Special attention has
Figure 1 shows several pulse sequences that we have exbeen focused on some heteronuclear gradient-based 1D perimentally tested. They are all designed to achieve selecschemes featuring selective polarization transfer from carbon tive polarization transfer from a specific carbon to its directly to proton using the basic HMQC and HSQC pulse trains attached proton via 1 J CH . In order to make an accurate com-(3-6), which can also be used as starting building blocks parative study, all experiments have been recorded using the in more sophisticated 1D sequences such as 1D 13 C-edited same experimental conditions in a Bruker ARX400 spec-TOCSY, NOESY, or ROESY experiments (4). Interesting trometer using an inverse broadband probehead incorporatadvantages of such gradient-based experiments are the lack ing a self-shielded Z-gradient coil. As a model, we have of need for a complete phase-cycle process, the perfect supchosen the H-1/C-1 pair ( 1 J CH Å 169.8 Hz) of a 0.2 M pression of the large unwanted 1 H-12 C magnetization withsucrose solution in D 2 O: out the need of the BIRD cluster, the minimization of dynamic range limitations, and the better tolerance to errors in parameter setup or to spectrometer instabilities. However, these gradient 1D experiments suffer from the theoretical loss of sensitivity by a factor of 2, compared to phase-cycled analogues, due to the selection by the gradients of only one of the two coherence-transfer pathways (CTPs) that can contribute to the detected signal. In practice, the use of a much higher setting of the receiver gain in PFG experiments partially overcomes these sensitivity drawbacks.
In the present work, the effectiveness of heteronuclear polarization transfer for several gradient-based 1D schemes is experimentally examined in terms of sensitivity. In addi-Four scans preceded by two dummy scans were acquired tion, new modified 1D pulse sequences offering improved for each 1D spectrum using a prescan delay of 1 s, an insensitivity for this heteronuclear polarization transfer are terpulse delay D (Å1/4 1 J CH ) of 1.5 ms, and an acquisition also proposed and discussed. These improved experiments recording time of 1.27 s (Fig. 2). The length of each gradient use the known principle of the preservation-of-equivalent-(shaped as a 5% truncated Gaussian) was 1 ms and the pathways (PEP) approach (7, 8), currently implemented in maximum strength was set at 20 G/cm. Selective 90Њ or many multidimensional heteronuclear experiments. This is 180Њ carbon pulses with a duration of 12 ms also had a Gaussian profile truncated to 5%. A WALTZ-16 pulse train was simultaneously applied to achieve better selectively during the application of these selective carbon pulses and to