The principal reason behind application of broadband proton decoupling of heteronuclei is removal of the coupling struc- ture to concentrate signal intensity, thereby improving the signal-to-noise ratio and reducing resonance overlap (Fig. 4.13).
Additional gains such as signal enhancements from the NOE and clarification of any remaining homonuclear couplings may also arise. Set against these obvious benefits is the loss of multiplicity information present in the proton-coupled spec- trum (Fig. 4.13a), meaning there is no way, a priori, of distinguishing, for example, a methine from a methylene carbon resonance. It is therefore desirable to be able to record fully proton-decoupled spectra yet still retain valuable multiplicity data. Some of the earliest multipulse sequences were designed to achieve this aim. These were based on simple spin-echoes and provided spectra in which the multiplicities were encoded as signal intensities and signs. Despite competition from 1D and 2D methods based on polarisation transfer described shortly, these techniques still find use in chemistry laboratories.
As they are also rather easy to understand, they provide a suitable introduction to the idea of spectrum multiplicity editing.
FIGURE 4.13 The carbon-13 spectrum of a-pinene 4.2 in CDCl3. Spectra were acquired (a) without proton irradiation at any stage (coupled spectrum without NOE), (b) with gated decoupling (coupled spectrum with NOE), (c) with inverse-gated decoupling (decoupled spectrum without NOE) and (d) pow- er-gated decoupling (decoupled spectrum with NOE). All other experimental conditions were identical and the same absolute scaling was used for each plot.
www.pdfgrip.com
One-Dimensional Techniques Chapter | 4 149
The sections that follow utilise the rotating-frame vector model to explain pictorially the operation of the experiments, and familiarity with the introduction in the Section 2.2 is assumed.
4.3.1 J-Modulated Spin-Echo
One of the simplest approaches to editing is the J-modulated spin-echo sequence [30] (Fig. 4.16a, also referred to as the spin-echo Fourier transform (SEFT, [31]) which can be readily appreciated with reference to the vector model.
The key to understanding this sequence is the realisation that the evolution of carbon magnetisation vectors under the influence of the 1JCH coupling only occurs when the proton decoupler is gated off, whilst at all other times only carbon chemical shifts are effective. Further simplification comes from ignoring carbon chemical shifts altogether since shift evolution during the first ∆ period is precisely refocused during the second by the 180 degree (C) refocusing pulse
FIGURE 4.15 Phosphorus decoupling of a proton spectrum. Simplification of the conventional proton spectrum (a) of the palladium phosphine 4.4 in CDCl3 is possible by application of broadband phosphorus decoupling (b). All long-range 1H–31P couplings are removed, as is most apparent for the alkene proton (7.2 ppm) and the ortho protons of the phenyl rings (above 7.8 ppm).
FIGURE 4.14 Application of selective proton decoupling in the measurement of heteronuclear long-range proton–carbon coupling constants.
Partial carbon-13 spectra are shown for (a) the CO groups and (b) the CN group. Lower traces are from the fully proton-coupled carbon-13 spectrum and the upper traces from that in which the methyl ester protons of 4.3 were selectively decoupled to reveal the three-bond coupling of the carbonyl carbon across the alkene.
www.pdfgrip.com
150 High-Resolution NMR Techniques in Organic Chemistry
(see Section 2.2). Thus, to understand this sequence one only need consider the influence of heteronuclear coupling during the first ∆ period.
Consider events with ∆ = 1/J s, for which the relevant vector evolution for different multiplicities is illustrated in Fig. 4.17. Since chemical shifts play no part, quaternary carbons remain stationary along +y during ∆ (long-range C─H coupling that may exist will be far smaller than the one-bond coupling and may be considered negligible). Doublet vectors for a methine pair evolve at ±J/2 Hz, so will each rotate through one-half cycle in 1/J s and hence meet once more along
−y. As these now have a 180 degrees phase difference with respect to the quaternary signals, they will ultimately appear inverted in the final spectrum. Applying these arguments to the other multiplicities shows that methylene vectors evolving
FIGURE 4.17 Evolution of carbon magnetisation vectors under the influence of proton–carbon couplings. Vectors are shown for carbon singlets (C), doublets (CH), triplets (CH2) and quartets (CH3).
FIGURE 4.16 J-modulated spin-echo sequences. (a) The decoupler-gated variant and (b) the pulsed variant.
www.pdfgrip.com
One-Dimensional Techniques Chapter | 4 151
at ±J Hz will align with +y whilst methyl vectors evolving at ±J/2 and ±3J/2 Hz will terminate along −y. More generally, by defining an angle u such that u = 180J∆ degrees, the signal intensities of carbon multiplicities I vary according to:
C: I = 1 CH: I ∝ cos u CH2: I ∝ cos2 u CH3: I ∝ cos3 u
as illustrated in Fig. 4.18. The spectrum for ∆ = 1/J s (u = 180 degrees) will therefore display quaternaries and methylenes positive with methines and methyls negative if phased as for the one-pulse carbon spectrum (Fig. 4.19b), although edited spectra are often presented with methine resonances positive and methylene negative, so check local convention. The car- bon multiplicities therefore become encoded as signal intensities and at least some of the multiplicity information lost in the single-pulse carbon experiment has been recovered, whilst still benefitting from decoupled resonances. The application of proton decoupling for all but a short time period ∆ also ensures spectra are acquired with enhancement from the NOE.
FIGURE 4.18 The variation of carbon signal intensities in the J-modulated spin-echo as a function of the evolution time ∆ (u = 180J∆ degrees).
FIGURE 4.19 Carbon spectrum of camphor 4.1 edited with the J-modulated spin-echo sequence. (a) Conventional carbon spectrum (carbonyl not shown), and edited spectra with (b) ∆= 1/J (u = 180 degrees) and (c) ∆ = 1/2J (u = 90 degrees) with J assumed to be 130 Hz. Some breakthrough of protonated carbons is observed in (c) due to variations in coupling constants within the molecule.
www.pdfgrip.com
152 High-Resolution NMR Techniques in Organic Chemistry
Setting ∆ = 1/2J s (q = 90 degrees) corresponds to a null for all protonated carbons (Fig. 4.18), so producing a quaternary- only spectrum (Fig. 4.19c). The accuracy of editing is clearly critically dependent on the correct setting of ∆, which is in turn dependent upon J, and as there are likely to be a wide variety of J values within a sample, one is forced to make some compromise setting for ∆. One-bond proton–carbon couplings range between 125 and 250 Hz, although they are more commonly between 130 and 170 Hz (Table 4.3), so a typical value for ∆ would be ≈7 ms (1JCH ≈ 140 Hz) when aromatics and carbon centres bearing electronegative heteroatoms are anticipated. If couplings fall far from the chosen value the cor- responding resonance can display unexpected and potentially confusing behaviour. Alkyne carbons are particularly prone to this, owing to exceptionally large 1JCH values.
The spin-echo experiment is particularly simple to set-up as it does not require proton pulses or their calibration, a desir- able property when the experiment was first introduced but of little consequence nowadays. The same results can, in fact, be obtained by the use of proton 180 degree pulses rather than gating of the decoupler [32] (Fig. 4.16b). In this case the
∆ period is broken in two periods of 1/2J separated by simultaneous application of proton and carbon 180 degrees pulses.
These serve to refocus carbon chemical shifts but at the same time allow couplings to continue to evolve during the second
∆/2 period (Section 2.2). Hence, the total evolution period in which coupling is active is ∆ as in the decoupler-gating experi- ment, and identical modulation patterns are produced. It is this shorter, pulsed form of the heteronuclear spin-echo that is widely used within numerous pulse sequences to refocus shift evolution yet leave couplings to evolve.
4.3.2 APT
The principle disadvantage of the J-modulated spin-echo described above is the use of a 90 degree carbon excitation pulse which, as discussed in Section 4.1, is not optimum for signal averaging and may lead to signal saturation, notably of quaternary centres. The preferred approach using an excitation pulse width somewhat less than 90 degrees requires a slight modification of the J-modulated experiment, giving rise to the attached proton test (APT) sequence [33] (Fig. 4.20). Use of a small tip angle excitation pulse leaves a component of magnetisation along the +z-axis, which is inverted by the 180 degree (C) pulse that fol- lows. In systems with slowly relaxing spins, this inverted component may cancel magnetisation arising from relaxation of the transverse components, leaving little or no net signal to observe on subsequent cycles. It is therefore necessary to add a further 180 degree carbon pulse which returns the problematic −z component to +z prior to acquisition, thereby eliminating possible cancellation. Transverse components also experience a 180 degrees rotation prior to detection, but are otherwise unaffected beyond this phase inversion. The APT experiment is commonly used with excitation angles of 45 degrees or less and is thus better suited to signal averaging than the basic echo sequence, but otherwise gives similar editing results.
The poor editing accuracy of spin-echoes in the presence of a wide range of J values and the inability to fully char- acterise all carbon multiplicities are the major limitations of these techniques. More complex variations on the pulsed J-modulated spin-echo are to be found that do allow complete decomposition of the carbon spectra into C, CH, CH2 and
FIGURE 4.20 The APT sequence. An excitation pulse (φ) less than 90 degrees may be used.
TABLE 4.3 Typical Ranges for One-Bond Proton–Carbon Coupling Constants
Carbon Environment Typical 1JCH Range (Hz)
Aliphatic, CHn- 125–135
Aliphatic, CHnX (X = N, O, S) 135–155
Alkene 155–170
Alkyne 240–250
Aromatic 155–165
www.pdfgrip.com
One-Dimensional Techniques Chapter | 4 153
CH3 sub-spectra [34] and which also show greater tolerances to variations in 1JCH [35]. Likewise, J-compensated APT sequences have been developed for greater tolerance to a spread of J values [36], and for the direct generation of complete sub-spectra [37]. Invariably, the simplest sequences still find widest use.