BR-24 pulse sequence

The H cramps NMR measurement was performed on a Chemagnetics CMX 300 spectrometer operating at 300.16 MHz, equipped with a 5 mm CRAMPS probe. The BR-24 pulse sequence"" was used, and 7r/2 pulse width was 1.3 ps. The rotational frequency was exactly controlled in the range 1.5 to 2.0 kHz, and the cycle time of BR-24 was 108 ps. The recycle delay was 10 s and spectra were usually accumulated 32 times to achieve a reasonable signal-to-noise ratio for the samples. The H chemical shift was calculated with a scaling factor of 0.40 for all samples, which was determined experimentally. The H cramps spectra were recorded first without internal standard, and calibrated afterwards with internal Si-rubber (5 = 0.12) relative to TMS (5 = 0). The typical half-width was 30 Hz, and the total measurement time for one sample was usually 5 min. 

BR-24 pulse sequence A 90° pulse width of 1.3 us is used for the BR-24 pulse sequence. The cycle time of the BR-24 is 108 /zs, corresponding to a t of 3 /zs. The MAS rotational frequency is controlled at 2.0 kHz. The H chemical shift has been calculated with a scaling factor of 0.40 for all samples, which was determined experimentally. 

Thus, it is concluded that (1) the quadrupolar N nuclei are responsible for NH signal broadening and the asymmetric line shape (2) the asymmetric line shape is different for a-helix and /3-sheet conformations and (3) the true NH proton chemical shift of a-helical [Ala ]n-2 can be determined (8.0 ppm) but that of [Ala ] - (/3-sheet) is difficult to determine because of the low S/N ratio using BR-24 pulse sequences at 2.0 kHz MAS speed. 

Figure 41 shows the H CRAMPS NMR spectra of a-helical and /3-sheet poly(L-leucines) using the BR-24 pulse sequence at 2.0 kHz MAS speed (A) [Leu]n-2 (natural abundance, a-helix), (B) [Leu ] -2 (99 al. /o N, a-helix), (C) [Leu]n-1 (natural abundance, /3-sheet) and (D) [Leu ]n-1 (99 at. /o N, /3-sheet) in the solid state. The H NMR spectra show high-resolution signals (H, H" and side-chain protons) of poly(L-leucines). The conformations of [Leu]n-2 and [Leu ]n-2, and [Leu]n-1 and [Leu ]n-1 are confirmed independently by the H" chemical shift (a-helix 4.0 ppm, -sheet 5.5 ppm). The peak around 4.0 ppm in Fig. 41(C) and (D) indicates that the samples take the a-helix component.

Figure 42 shows the H CRAMPS NMR spectra of N-labelled poly(L-alanines) in the a-helical form (A) BR-24 pulse sequence at 2.0 kHz MAS speed, (B) MREV-8 pulse sequence at 3.5 kHz MAS speed in the solid state. The intensity of the NH proton signal of poly(L-alanine) was increased when measuring with the MREV-8 pulse sequence at a faster MAS speed (3.5 kHz). 

Our procedure also offers the opportunity to study the effects of various kinds of pulse errors. The general result of our simulations is that known line-narrowing m.p. sequences like the MREV-8 (Mansfield, 1971 Rhim et al., 1973a, b) and BR-24 (Burum et ah, 1979b) sequences perform so well and are so robust against pulse errors that the spectral resolution in actual experiments will rarely be limited by the m.p. sequence as such provided that T can be made as short as 1.5 (MREV-8) or 3 /is (BR-24). lliese are numbers that are well within the reach of our spectrometer and other specialized m.p. spectrometers. 

Fio. 6. Dependence of the residual dipolar width of the line ascribed to proton 1 of the model system versus the pulse spacing t for the MREV and BR-24 sequences. 

When finite pulsewidths are taken into account, second-order dipolar terms do not drop out completely for the BR-24 sequence and we learn from the tp = t data in Fig. 6 that such terms dominate the residual linewidth. The resolution obtained with the MREV sequence for = r is inferior by a factor of 2.4 compared to that for tp => 0, irrespective of the pulse spacing r. By contrast, the gap between the tp = t and tp 0 data widens for BR-24 when r becomes smaller. 

Stipulating a phase error of +1° of the +y pulses in the BR-24 sequence Up = 0.75 /AS, r = 3 /AS, /3 = 90°) results in the spectrum shown in Fig. 8b. For comparison we have reproduced the r = 3-/as BR-24 spectrum of Fig. 3 in Fig. 8a. We recognize that a (small) phase error hardly affects the widths of the lines its major consequence is a common shift of all lines of about 0.5 ppm toward higher frequencies. For resonances near zero... 

The scaling factor of a m.p. sequence depends on p (Haeberlen, 1976). A variation of p during the sequence therefore causes a chirp of each resonance. Our simulation program allows us to quantify this effect also. In Fig. 10 we show a simulated BR-24 spectrum of our model system that assumes an exponential power droop that amounts to no more than a 1% decrease of p after 100 BR-24 cycles, that is, after 2400 pulses. Note the asymmetry of the lines and the wiggles at their feet that are indicative of the chirp. In Section IV we present experimental m.p. spectra that display exactly these features. 

On the 1-ppm resolution level, both the MREV and BR-24 sequences are robust against pulse errors provided that the ratio can be kept smaller than about 3/8. This statement emphasizes again the importance of being able to flip the proton magnetization through 90° in less than 1 /is. 

Spectrometer does not fall far behind the theoretical limitations, if it does at all. The quality of the pulses (uniformity and constancy of the flip angles and the rf phases) is sufficiently high that pulse errors hardly play a role as a resolution-limiting factor. The tightest theoretical limitation is the necessarily finite width of the rf pulses, which is particularly acute for the BR-24 sequence. The next significant step to enhance the resolution in solid state proton m.p. spectroscopy may well require either 90° pulses shorter than, say, 500 ns (this would be the brute force method) or another clever idea. 

The scaling factor of the BR-24 or MREV-8 sequences is a little smaller than the theoretical value (MREV-8 0.471 BR-24 0.385) for ideal square pulses. The scaling factor depends on the pulse length, the cycle time, the pulse sequence, and properties of the machine. However, it is impossible to make a... 

An MAS frequency of 1.5-2.0 kHz is generally used with the BR-24 to prevent interference between the MAS and the multiple-pulse sequence. MAS frequency of 5-6 kHz is possible with MREV-8. 

Solid-state H CRAMPS NMR spectra were measured with a Chemagnetics CMX 300 spectrometer operating at 300 MHz. The internal standard was Si-rubber ((5 = 0.12) relative to (CH3)4Si (6 = 0). BR-24 and MREV-8 pulse sequences were applied. The radio frequency (RF) powers and duration windows (r) of these pulse sequences were adjusted so as to obtain the best resolution for adipic acid. 

From the H CRAMPS NMR spectra, therefore, it was possible to determine the NH proton chemical shift value for [Ala ]n-2 (a-helix 6 = 8.0) which is identical with that determined using BR-24 (2.0 kHz). Further, it was possible to determine the " NH proton chemical shift for [Ala ]n-1 (/3-sheet, 6 = 8.6) using the MREV-8 pulse sequence at 3.5 kHz. However, unfortunately, the NH proton chemical shift values for [Ala]n-2 and [Ala]n-t could not be determined because the line shapes of the " NH signals exhibit an asymmetric doublet pattern in this system also. Thus, it is found that determination of the true NH chemical shift of poly(L-alanines) can be achieved to measure fully N-labelled samples at higher MAS speed (3.5 kHz) and that these chemical shifts depend on conformation (a-helix 6 = 8.0 /3-sheet S = 8.6). This is the first determination of the true NH proton chemical shifts of poly(L-alanines) by H CRAMPS NMR. 

Fig. 15. Four multi-pulse sequences for suppressing homonuclear dipolar interaction (a) WAHUHA, (b) MREV-8, (c) BR-24 and (d) CORY-24. One cycle is drawn for each sequence, l e longer delays are double the length of the shorter ones.

The basic principle behind the multiple-pulse NMR techniques to achieve line narrowing (i.e., eliminate the H- H dipolar interaction) is to manipulate the H spin system with r.f. pulses rather than by motion of the whole system, as is done with MAS. This manipulation is performed by using a series of well-timed r.f. pulses such that the average Hamiltonian over the entire period of the pulse sequence does not include the homonuclear dipolar interaction, but still maintains a scaled-down chemical shift e ct. Because of the strict requirements on r.f. pulse widths, shapes, phasing and timing, the multiple-pulse techniques represent some of the most difficult solid-state NMR techniques to implement on a routine basis. The most popular multiple-pulse techniques are currently the eight-pulse MREV-8 and the 24-pulse BR-24 sequence. ... 

The original homonuclear line-narrowing pulse sequence (WAHUHA) [23] was a four-pulse sequence later elaborations involve more pulses in the total cycle and offer compensation for pulse imperfections and/or higher order averaging of the homonuclear dipolar interaction [24]. Currently, the most popular multiple pulse sequences are the eight-pulse MREV-8 sequence [24a] and the 24-pulse BR-24 sequence [24b]. 

---