CHAPTER 8 A, Introduction ‘The perennial student question: Where do we start? ‘The instructor will be sympathetic but not rigidly pre- scriptive. There are, however, guidelines that do start with the prescriptive statement: Go for the molecular formula, Why? Simply because it is the single most use~ ful bit of information available to the chemist and is worth the effort sometimes necessary. It provides an overall impression of the molecule (ie., the number and kinds of atoms), and it provides the index of hydrogen deficiency —in other words, the sum of the number of rings and of double and triple bonds (Section 2.6). Development of the molecular formula starts with recognition of the molecular ion peak (Section 2.5). We assume the usual situation: High-resolution MS instru- mentation is not readily available, Let us also assume for now that the peak of highest m/z (except for its iso- tope peaks) is the molecular ion peak and is intense enough so that the isotope peak intensities can be de- termined accurately, and the presence and number of S, Br, and Cl atoms can be ascertained. If the molecular ion peak is an odd number, an odd number of N atoms is present. A search of the infrared spectrum for the familiar characteristic groups is now in order. Note in particular the unsaturated functional groups. Look also at the frag- mentation pattern of the mass spectrum for recogniz- able fragments, With this information in hand, search the proton NMR spectrum for confirmation and further leads. Se- lect an entry point; pethaps a CH;C=O or an OCH,CH, moiety as in Problem 8.1. If the spectrum is, first order, or at least resembling one, determine the total proton count and ratios of groups of chemical shift—equivalent protons from the integration, Look at the C/DEPT spectra; determine the car- bon and proton counts and the numbers of CHs, CH; and CH groups. Overlap of proton absorptions is com- mon, but absolute coincidence of nonequivalent "C peaks is quite rare with a high-resolution instrument. Now, select the most likely molecular formula from 301 Appendix A of Chapter 2, and determine the index of hydrogen deficiency. In addition to difficulties caused by unresolved or overlapping peaks, discrepancies may appear between the selected molecular formula and the 'H and '8C counts because of the presence of elements of symmetry. But this information also contributes to an understanding of the molecular structure (Section 5.23) Difficulty often starts with uncertainty in the cho of a molecular ion peak. Many laboratories use chemical ionization as a routine supplement to electron impact, and of course, access to a high-resolution instrument desirable for more difficult problems. Students are urged to develop their own ap- proaches. To provide practice in the use of the newer techniques, we have sometimes presented more infor- mation than needed, but other Problems should provide compensatory frustration to simulate the real world. Remember the overall strategy: Play the spectra against one another, focusing on the more obvious features. De- velop a hypothesis from one spectrum; look to the other spectra for confirmation or denial; modify the hypoth- esis if necessary. The effect is synergistic, the total infor- mation being greater than the sum of the individual parts. With the high resolution now available, many NMR spectra are first order, or nearly so, and can be inter- preted by inspection with the leads furnished by the ‘mass and infrared spectra. Nevertheless, a rereading of Sections 4.12 through 4.16 may engender caution. ‘As an example, consider two similar compounds: H He one oe H H H H ae Both rings exist as rapidly flexing ring conforma tions, but only in Compound A do the protons of each302 Chapter’8 A. Introduction CH, group interchange to become chemical-shift equiv- alent (enantiotopes) (see Section 4.12.3.3). Only Com- pound A has a plane of symmetry in the plane of page through which the protons interchange. From left to right in the spectrum, we predict for Compound A: H-5, a two-proton triplet; H-3, a two- proton triplet; H-4, a two-proton quintet (assuming nearly equal coupling constants). Given modest reso- lution, the spectrum is first order. Compound B has no symmetry element in the pla- nar conformation. C-5 isa chiral center, and the protons of each CH, group are diastereotopic pairs. Each proton of the pair has its own chemical shift. The H-4 protons adjacent to the chiral center are distinetly separated, but the H-3 protons are not, at 300 MHz. Each proton of a diastereotopic pair couples geminally with the other and independently (different coupling constants) with the vicinal protons to give complex multiplets. The possibility of a chiral center should always be kept in mind (see Problem 8.3); toujours la stéréochimie. ‘Compounds A and B are assigned Problems in Chap- ter 9, ‘The power of 2-D spectra will become more evident as we work through the problems in Chapters 8 and 9. Itis often not necessary to examine all of the spectra in detail before proposing—tentatively—possible struc- tures or fragments. Spectral features predicted for the postulated structures or fragments are compared with the observed spectra, and structural modifications are made to accommodate discrepancies. ‘These suggestions are illustrated by the following solved Problems presented in increasing order of diffi- culty. The assigned Problems of Chapter 9, again in in- creasing order of difficulty, will provide the essential practice, Most students enjoy problem solving and rise to the References challenge. They also begin to appreciate the elegance of chemical structure as they interpret spectra. Good sleuthing! Be wary of chirality, diastereotopes, virtual coupling, dihedral angles of about 90°, and magnetic nonequivalence. Finally, what are the requirements for proof of structure? Ultimately, it is congruence of all available spectra with those of a pure, authentic sample obtained under the same conditions and on the same instruments. Obviously, some compromises are acceptable. Con- gruence with published spectra or spectral data is con- sidered acceptable for publication, but this cannot apply to a new compound, which must then be synthesized. ‘Computer programs for simulation of proton NMR. spectra are available. If accurate measurements of chemical shifts and coupling constants for all of the pro- tons can be obtained, the simulated spectrum will be congruent with the actual spectrum. In many cases, at least some of the spin systems will be first order. If not, reasonable estimates of shifts and coupling constants may be made, and the iterative computer program will adjust the values until the simulation matches the actual spectrum —assuming, of course, that the identification is valid.* Unless otherwise labeled here and in Chapter 9, the NMR spectra were obtained at 300 MHz for protons and 75.5 MHz for 8C; CDCl, was the solvent unless otherwise labeled. The IR spectra were obtained neat (ie., no solvent) unless otherwise labeled. The mass spectra were obtained by GC/MS. The COSY spectra are DOF-COSY spectra. The labeled frequency for all 2-D NMR spectra is that of the acquired signal (F2) The following Problem sets are available for further practice. Atta-ur-Rahman, and Choudhary, M.1. (1996). Solving Prob: lems with NMR. New York: Academic Press. Bates, R.B,,and Beavers, W.A. (1981). Carbon-13 NMR Spec tral Problems. Clifton, NJ: Humana Press. Braun S., et al, (eds.). (1996). 100 and More Basic NMR Ex- periments, New York: VCH Breitmaier, E. (1993). Structure Elucidation by NMR in Or- ganic Chemistry, A Practical Guide. New York: Wiley. Davis, R., and Wells, C.H.J. (1984). Spectral Problems in Or- ‘ganic Chemistry. New York: Chapman and Hall Duddeck, H., and Dietrich, W. (1992). Structural Elucida: tion by Modern NMR. A Workbook, 2nd ed. New York: Springer-Verlag Field, L.D., Sternhell, S., and Kalman, J.R. (1995). Organic Structures from Spectra, New York: Wiley. Fuchs, P.L., and Bunnell, C.A. (1979). Carbon-13 NMR Based Organic Spectral Problems. New York: Wiley. Sanders, J.K.M., Constable, E.C., and Hunter, B.K. (1989). Modern NMR Spectroscopy; A Workbook of Chemical Problems. Oxford: Oxtord University Press. * Spectra ean be simulated on the computer of a modern NMR spec trometer or on a PC. For example, see the Win-Daisy program, avail able from Bruker Instruments Incorp.. Billerica, Mass,CHAPTER 8 B. Solved Problems Compound 8.1 We start by gathering information in order to establish a molecular formula. We assume that the weak peak at m/z 144 is the molecular ion peak. It is so small that the intensities of its isotope peaks cannot be accurately measured. Since m/z 144 is an even number, there are . Natoms present. To begin, we tentatively assume that are no N, S, or halogen atoms present; this posture, of course, is quite shaky. From left to right, the proton integrator in the 'H NMR spectrum reads: 2, 2, 2, 3, 3—calibrated against the presumed methyl singlet at 2.17. From high to low frequency the "°C and DEPT spectra read: C, C, CH2, CH,, CH, CH,, CH. Thus, there are 12 protons and 7 carbon atoms in the molecular formula. Note that the DEPT CH subspectrum is omitted since there are no CH groups. The most likely molecular formula under unit mass 144 is C;H,O, (Chapter 2, Appendix A). The index of hydrogen deficiency is 2, and this should be immediately explored. ‘The IR spectrum shows a strong, broad C=O peak at about ~1725 em-', which accounts for one unsatu- rated site and for one O atom. The "C spectrum shows a ketone C=O group at ~6 208, and an ester C=O group at ~8 172.5; the latter assignment is reinforced by typical ethyl ester peaks in the IR spectrum at ~1160.cm~ and ~1030cm-'. The broad peak at ~1725 cm must represent both C=O groups. The three O atoms in the molecular formula are ac- counted for. With this information in hand, interpretation of the 1H spectrum is straightforward. The methyl singlet men- tioned above must be attached to the ketone C=O group to give us one end of the molecule, CH;—C —, ° which also accounts for the base peak in the mass spectrum at m/z 43. The three-proton triplet and the strongly deshielded, two-proton quartet account for the — memes CH, moiety at the other end of ° the molecule, Confirmation is provided by the strong peak at m/z 99 (characteristic loss of O—CH,—CH;). ‘The NMR spin systems are Ag, aX, and AgX>. Filling in between the two ends of the molecule re- quires little imagination. All that remain in the 'H spec- trum are two two-proton triplets—surely two adjacent CH; groups. Hence: cH ¢ —CH,—CH;— fi —O—CH,—CH, | oO ° Ethyl levulinate, Ethyl 4-oxopentanoate Let us return for a moment to the mass spectrum: Note that the loss of 15 units (loss of CH;) to give a moderate peak of m/z 129 provides confirmation that the weak peak at m/z 144 is indeed the molecular ion peak. Loss of 45 units to give the strong peak at m/z 99 provides further confirmation, ‘Assignment of the shifts of the CH, groups adjacent to the C=O groups is ambiguous. Assignment can be achieved by obtaining an HMBC spectrum (Chapter 6) which would show correlation (long-range coupling) be- tween the groups adjacent to the ketone C=O group (Chapter 6). For further discussion, consider and reject the fol- lowing isomers of ethyl levulinate: CH,—CH,— ¢ —CH,— f —O—CH,—CH, ° ° HE ‘CH,—CH,—CH,— i —O—CH,—CH, oO oO CH)—CH,—CH,— E— €—O—CH—CHs H,—CH,— FC CHF —O—CH, CH, —O— € CH, CH, oO H CH—CH, ro » 5 cH—cH,—0—-H,c% “o® So