(no neighbors) (one neighbor) (two neighbors) (three neighbors) (four neighbors) (five neighbors) (six neighbors)

J2 J2

J2 J2

Singlet Doublet

Triplet Doubledoublet

Quartet Quintet



Figure 1.5

The strength of the spin-spin splitting interaction, measured by the peak separation ("J value") in units of hertz, depends in a predictable way on the dihedral angle defined by H a-C-C-Hb, so that information can be obtained about the stereochemistry and conformation of molecules in solution. Because of this dependence on the geometry of the interceding bonds, it is possible to have couplings for two neighbors with different values of the coupling constant, J. This gives rise to a splitting pattern with four peaks of equal intensity: a double doublet (Fig. 1.5).

A third type of information available from NMR comes from the nuclear Overhauser enhancement or NOE. This is a direct through-space interaction of two nuclei. Irradiation of one nucleus with a weak radio frequency signal at its resonant frequency will equalize the populations in its two energy levels. This perturbation of population levels disturbs the populations of nearby nuclei so as to enhance the intensity of absorbance at the resonant frequency of the nearby nuclei. This effect depends only on the distance between the two nuclei, even if they are far apart in the bonding network, and varies in intensity as the inverse sixth power of the distance. Generally the NOE can only be detected between protons (1H nuclei) that are separated by 5 A or less in distance. These measured distances are used to determine accurate three-dimensional structures of proteins and nucleic acids.

1.1.7 Pulsed Fourier Transform (FT) NMR

Early NMR spectrometers recorded a spectrum by slowly changing the frequency of a radio frequency signal fed into a coil near the sample. During this gradual "sweep" of frequencies the absorption of energy by the sample was recorded by a pen in a chart recorder. When the frequency passed through a resonant frequency for a particular nucleus in the sample, the pen went up and recorded a "peak" in the spectrum. This type of spectrometer, now obsolete, is called "continuous wave" or CW. Modern NMR spectrometers operate in the "pulsed Fourier-transform" (FT) mode, permitting the entire spectrum to be recorded in 2-3 s rather than the slow (5 min) frequency sweep. The collection of nuclei (sample) is given a strong radio frequency pulse that aligns the nuclei so that they precess in unison, each pointing in the same direction at the same time. The individual magnetic fields of the nuclei add together to give a measurable rotating magnetic field that induces an electrical voltage in a coil placed next to the sample. Over a period of a second or two the individual nuclei get out of synch and the macroscopic signal dies down. This "echo" of the pulse, observed in the coil, is called the free induction decay (FID), and it contains all of the resonant frequencies of the sample nuclei combined in one cacophonous reply. These data are digitised, and a

Figure 1.6

computer performs a Fast Fourier Transform to convert it from an FID signal as a function of time (time domain) to a plot of intensity as a function of frequency (frequency domain). The "spectrum" has one peak for each resonant frequency in the sample. The real advantage of the pulsed-FT method is that, because the data is recorded so rapidly, the process of pulse excitation and recording the FID can be repeated many times, each time adding the FID data to a sum stored in the computer (Fig. 1.6). The signal intensity increases in direct proportion to the number of repeats or "transients" (1.01,2.01,2.99,4.00), but the random noise tends to cancel because it can be either negative or positive, resulting in a noise level proportional to the square root of the number of transients (0.101, 0.145, 0.174, 0.198). Thus the signal-to-noise ratio increases with the square root of the number of transients (10.0, 13.9, 17.2, 20.2). This signal-averaging process results in a vastly improved sensitivity compared to the old frequency sweep method.

The pulsed Fourier transform process is analogous to playing a chord on the piano and recording the signal from the decaying sound coming out of a microphone (Fig. 1.7). The chord consists of three separate notes: the "C" note is the lowest frequency, the "G" note is the highest frequency, and the "E" note is in the middle. Each of these pure frequencies gives a decaying pure sine wave in the microphone, and the combined signal of three frequencies is a complex decaying signal. This time domain signal ("FID") contains all three of the frequencies of the piano chord. Fourier transform will then convert the data to a "spectrum"—a graph of signal intensity as a function of frequency, revealing the three frequencies of the chord as well as their relative intensities. The Fourier transform allows us to record all of the signals simultaneously and then "sort out" the individual frequencies later.

Audio Figure 1.8

1.1.8 NMR Hardware

An NMR spectrometer consists of a superconducting magnet, a probe, a radio transmitter, a radio receiver, an analog-to-digital converter (ADC), and a computer (Fig. 1.8). The magnet consists of a closed loop ("solenoid") of superconducting Nb/Ti alloy wire immersed in a bath of liquid helium (bp 4 K). A large current flows effortlessly around the loop, creating a strong continuous magnetic field with no external power supply. The helium can ("dewar") is insulated with a vacuum jacket and further cooled by an outer dewar of liquid nitrogen (bp 77 K). The probe is basically a coil of wire positioned around the sample that alternately transmits and receives radio frequency signals. The computer directs the transmitter to send a high-power and very short duration pulse of radio frequency to the probe coil. Immediately after the pulse, the weak signal (FID) received by the probe coil is amplified, converted to an audio frequency signal, and sampled at regular intervals of time by the ADC to produce a digital FID signal, which is really just a list of numbers. The computer determines the timing and intensity of pulses output by the transmitter and receives and processes the digital information supplied by the ADC. After the computer performs the Fourier transform, the resulting spectrum can be displayed on the computer monitor and plotted on paper with a digital plotter. The cost of an NMR instrument is on the order of $120,000-$5,000,000, depending on the strength of the magnetic field (200-900 MHz proton frequency).

1.1.9 Overview of 1H and 13C Chemical Shifts

A general understanding of the trends of chemical shifts is essential for the interpretation of NMR spectra. The chemical shifts of 1H and 13 C signals are affected by the proximity of electronegative atoms (O, N, Cl, etc.) in the bonding network and by the proximity to unsaturated groups (C=C, C=O, aromatic) directly through space. Electronegative groups shift resonances to the left (higher resonant frequency or "downfield"), whereas unsatu-rated groups shift to the left (downfield) when the affected nucleus is in the plane of the unsaturation, but have the opposite effect (shift to the right or "upfield") in regions above and below this plane. Although the range of chemical shifts in parts per million is much larger for 13 C than for 1H (0-220 ppm vs. 0-13 ppm), there is a rough correlation between the shift of a proton and the shift of the carbon it is attached to (Fig. 1.9). For a "hydrocarbon" environment with no electronegative atoms or unsaturated groups nearby, the shift is

Figure 1.9

near the upfield (right) edge of the range, with a small downfield shift for each substitution: CH > CH2 > CH3 (1H: 1.6,1.2,0.8; 13C: 30,20,10ppm). Oxygen has a stronger downfield-shifting effect than nitrogen due to its greater electronegativity: 3-4 ppm (1H) and 50-85 (13C) for CH-O. As with the hydrocarbon environment, the same downfield shifts are seen for increasing substitution: Cq-O (quaternary) > CH-O > CH2O > CH3O (13C around 85, 75, 65, and 55 ppm, respectively). Proximity to an unsaturated group usually is down-field shifting because the affected atom is normally in the plane of the unsaturation: CH3 attached to C=O moves downfield to 30 (13C) and 2.1 ppm (1H), whereas in HC=C (closer to the unsaturation) 13 C moves to 120-130 ppm and 1H to 5-6 ppm. The combination of unsaturation and electronegativity is seen in H-C=O: 190 ppm 13 C and 10 ppm 1H. There are some departures from this correlation of 1H and 13C shifts. Aromatic protons typically fall in the 7-8 ppm range rather than the 5-6 ppm range for olefinic (HC=C for an isolated C=C bond) protons, whereas 13C shifts are about the same for aromatic or olefinic carbons. Because carbon has more than one bond, it is sensitive to distortion of its bond angles by the steric environment around it, with steric crowding usually leading to downfield shifts. Hydrogen has no such effect because it has only one bond, but it is more sensitive than carbon to the through-space effect of unsaturations. For example, converting an alcohol (CH-OH) to an ester (CH-OC(O)R) shifts the 1H of the CH group downfield by 0.5 to 1 ppm, but has little effect on the 13 C shift.

1.1.10 Equivalence in NMR

Nuclei can be equivalent (have the same chemical shift) by symmetry within a molecule (e.g., the two methyl carbons in acetone, CH3COCH3), or by rapid rotation around single bonds (e.g., the three methyl protons in acetic acid, CH3CO2H). The intensity (integrated peak area or integral) of signals is directly proportional to the number of equivalent nuclei represented by that peak. For example, a CH3 peak in a molecule would have three times the integrated peak area of a CH peak in the same molecule.

1.1.11 Proton Spectrum Example

The first step in learning to interpret NMR spectra is to learn how to predict them from a known chemical structure. An example of a (proton) NMR spectrum is shown for

Figure 1.10

4-isopropylacetophenone (Fig. 1.10). The two isopropyl methyl groups are equivalent by symmetry, and each methyl group has three protons made equivalent by rapid rotation about the C-C bond. This makes all six Ha protons equivalent. Because they are far from any electronegative atom, these protons have a chemical shift typical of an isolated CH3 group: 0.8 ppm (see Fig. 1.9). The absorbance is split into two peaks (a doublet) by the single neighboring Hb proton. The six Ha protons do not split each other because they are equivalent. The integrated area of the doublet is 6.0 because there are six Ha protons in the molecule. The Hb proton is split by all six of the Ha protons, so its absorbance shows up as a septet (seven peaks with intensity ratio 1:6:15:20:15:6:1). Its integrated area is 1.0, and its chemical shift is downfield of an isolated CH2 (1.2 ppm) because of its proximity to the unsaturated aromatic ring (close to the plane of the aromatic ring so the effect is a downfield shift). The He methyl group protons are all equivalent due to rapid rotation of the CH3 group, and their chemical shift is typical for a methyl group adjacent to the unsaturated C=O group (2.1 ppm). There are no neighboring protons (the Hd proton is five bonds away from it, and the maximum distance for splitting is three bonds) so the absorbance appears as a single peak ("singlet") with an integrated area of 3.0. The Hc and Hd protons on the aromatic ring appear at a chemical shift typical for protons bound directly to an aromatic ring, with the Hd protons shifted further downfield by proximity to the unsaturated C=O group. Each pair of aromatic protons is equivalent due to the symmetry of the aromatic ring. The Hc absorbance is split into a doublet by the neighboring Hd proton (note that from the point of view of either of the Hc protons, only one of the Hd protons is close enough to cause splitting), and the Hd absorbance is split in the same way. Note that the J value (separation of split peaks) is the same for the Hc and Hd doublets, but slightly different for the Ha-Hb splitting. In this way we know, for example, that Ha is not split by either Hc or Hd.

1.1.12 Carbon Spectrum Example

The 13C spectrum of the same compound is diagramed in Figure 1.11. Several differences can be seen in comparison with the 1H spectrum. First, there is no spin-spin splitting due

to adjacent carbons. This is because of the low natural abundance of 13 C, which is only 1.1%. Thus the probability of a 13C occurring next to another 13C is very low, and splitting is not observed because 12C has no magnetic properties. Second, there is no spin-spin splitting due to the protons attached to each carbon. This is prevented intentionally by a process called decoupling, in which all the protons in the molecule are simultaneously irradiated with continuous low-power radio frequency energy at the proton resonance frequency. This causes each proton to flip rapidly between the upper and lower (disaligned and aligned) energy states, so that the 13C nucleus sees only the average of the two states and appears as a singlet, regardless of the number of attached protons. The lack of any spin-spin splitting in decoupled 13 C spectra means that each carbon always appears as a singlet. The multiplicity (s, d, t, q) indicated for each carbon in the diagram is observed only with the decoupler turned off and is not shown in the spectrum. Third, the peaks are not integrated because the peak area does not indicate the number of carbon atoms accurately. This is because 13C nuclei relax more slowly than protons, so that unless a very long relaxation delay between repetitive pulses is used, the population difference between the two energy states of 13C is not reestablished before the next pulse arrives. Quaternary carbons, which have no attached protons, relax particularly slowly and thus show up with very low intensity.

The molecular symmetry, indicated by a dotted line (Fig. 1.11) where the mirror plane intersects the plane of the paper, makes the two isopropyl methyl carbons Ca equivalent. Their chemical shift is a bit downfield of an isolated methyl group due to the steric crowding of the isopropyl group. Unlike protons,13C nuclei are sensitive to the degree of substitution or branching in the immediate vicinity, generally being shifted downfield by increased branching. Cb is shifted further downfield because of direct substitution (it is attached to three other carbons) and proximity to the aromatic ring. Ch is in a relatively uncrowded environment, but is shifted downfield by proximity to the unsaturated and electronegative carbonyl group. With the decoupler turned off, CH3 carbons appear as quartets because of the three neighboring protons. The aromatic CH carbons Cd and Ce are in nearly identical environments typical of aromatic carbons, and each resonance peak represents two carbons due to molecular symmetry. With the decoupler turned off, these peaks turn into doublets due to the presence of a single attached proton. The two quaternary aromatic carbons Cc and Cf are shifted further downfield by greater direct substitution (they are attached to three other carbons) and by steric crowding (greater remote substitution) in the case of Cc and proximity to a carbonyl group in the case of Cf. The chemical shift of the carbonyl carbon Cg is typical for a ketone. All three of the quaternary carbons Cc, Cf, and Cg have low peak intensities due to slow relaxation (reestablishment of population difference) in the absence of directly attached protons.


A few real-world examples will illustrate the use of XH and 13C chemical shifts and J couplings, as well as introduce some advanced methods we will use later. Two typical classes of complex organic molecules will be introduced here to familiarize the reader with the elements of structural organic chemistry that are important in NMR and how they translate into NMR spectra. Terpenoids are typical of natural products; they are relatively nonpolar (water insoluble) molecules with a considerable amount of "hydrocarbon" part and only a few functional groups—olefin, alcohol, ketone—in a rigid structure. Oligosaccharides are polar (water soluble) molecules in which every carbon is functionalized with oxygen— alcohol, ketone, or aldehyde oxidation states—and relatively rigid rings are connected with flexible linkages. In both cases, rigid cyclohexane-chair ring structures are ideal for NMR because they allow us to use J-coupling values to determine stereochemical relationships of protons (cis and trans). The molecules introduced here will be used throughout the book to illustrate the results of the NMR experiments.

1.2.1 Oligosaccharides

A typical monosaccharide (single carbohydrate building block) is a five or six carbon molecule with one of the carbons in the aldehyde or ketone oxidation state (the "anomeric" carbon) and the rest in the alcohol oxidation state (CH(OH) or CH2OH). Thus the anomeric carbon is unique within the molecule because it has two bonds to oxygen whereas all of the other carbons have only one bond to oxygen. Normally the open-chain monosaccharide will form a five- or six-membered ring as a result of the addition of one of the alcohol groups (usually the second to last in the chain) to the ketone or aldehyde, changing the C=O double bond to an OH group.

The six-membered ring of glucose prefers the chair conformation shown in Figure 1.12, with nearly all of the OH groups arranged in the equatorial positions (sticking out and roughly in the plane of the ring) with the less bulky H atoms in the axial positions (pointing up or down, above or below the plane of the ring). This limits the dihedral angles between neighboring protons (vicinal or three-bond relationships) to three categories: axial-axial (trans): 180° dihedral angle, large J coupling (~10 Hz); axial-equatorial (cis): 60° dihedral angle, small J (~4 Hz); and equatorial-equatorial (trans): 60° dihedral angle, small J (~4 Hz).

0 0

Post a comment