You are on page 1of 64
Chapter 10 THE THEORY OF PERICYCLIC REACTIONS In Chapters 10 and 1 we consider pericyelic reactions, processes characterized by bonding changes taking place through reorganization of electron pairs within a closed loop of interacting orbitals. ‘The bonding changes must be concerted in order for reaction to fit the pericyclic category; that is, all bonds breaking and forming must do so simultancously rather than in two or more steps. We shall use the term stepwise to imply a nonpericyclic pathway in which one or more bonds form (or break) in a first step to yield an intermediate that subsequently reacts by formation (breakage) of other bonds. The requirement for simultaneous bond formation and cleavage in a reaction does not mean that all changes must have taken place to the same ex stages along the reaction coordinate. So long as the changes are occurring together, it is possible for bond formation or cleavage at one site to run substantially ahead of bond formation or cleavage at another. A concerted reaction is therefore interpreted as ultaneous, but not necessarily synchronous, bond formation and breakage Pericyclic reactions represented for many years a difficult mechanistic problem because the apparent absence of intermediates left few concrete features that could be subjected to experimental study. Application of some fundamental principles of or- bital theory, initiated in 1965 by Woodward and Hoffmann! and since developed pericyclic tent at all ‘Woodward, R. B.: Hoff RJ. Am. Chem, Soe. 1965, 87, 395, 839 840 ‘Tue THrory oF PeRicyciic Reactions extensively by them® and by others,” have provided new insight into these reactions and have opened a new field of experimental investigation. This chapter considers the theoretical aspects, and Chapter 11 will take up applications and examples. 10.1 DEFINITIONS Woodward and Hoffmann* have subdivided pericyclic transformations into five cate- gories: cycloaddition, electrocyclic, sigmatropic, cheletropic, and group transfer reac- tions. Although one can, from a more generalized point of view, regard all these types as cycloadditions, it nient to follow this subdivision for purposes of thinking about both theoretical aspects and particular reactions. In this section we shall define and illustrate each class in order to arrive at an overall view of the kinds of questions: with which pericyclic reaction theory is concerned. is conv Cycloaddition Cycloaddition is a process in which two or more molecules condense to form a ring by transferring electrons from 7 bonds to new o bonds.” Some typical examples, charac- ed by the number of 7 electrons in the reacting components, are illustrated in ns 10.1-10.3.° The bonding changes are sometimes emphasized by the use of curved arrow: Equation 10.4. Although this notation serves as a useful book- keeping device to record the changes in bonding, one must nevertheless be careful when using it for pericyclic reactions not to attach literal significance to the arrows nor to their direction. As we shall sce in more detail later, the bonding changes occur through a set of orbitals interacting on the equivalent of a circular path, but a drawing like 1 must not be taken to imply that particular electron pairs are to be found at particular places on the circle or that pairs circulate in a particular direction, 2(a) Hoffmann, R.s Wooslward, R. B. J. Aim, Chem. Soe. 1965, 87, 2046; (b) Woodward, R. B.; Hoffimann, RJ Am. Chem. Se wd, RB. Ace, Chem. Res, 1968, 1, 17; (l) Woodward, R. B; Hoffinan, R. ‘Angew. Chem, Int, Ed. Engl, 1969, 8, 781; Verlag Chemie: Weinheim/Bergst., Geriany, aud Academic Press: New York, 1970, 3(a) Longuet-Higgins, H, Gz Abrahamson, E. W. J. Am. Chem, Soc. 1965, 87, 2045; (b) Zimmerman, HE. JA, Chien. Soe 1966, 88, 1564, 1566; Acc. Chem, Kes, 1971, 4, 272; Ace, Chem. Res. 1972, 5, 395: (e) Dewar, M.}.S. Teraliedron Suppl. 1966, Angew. Chem. Int. Ed. Bgl. U9TH, 10, 761; (A) Fukui, K. Ace, Chem Res. 1971, 4,57: (e) Fukui, K. Angew Chem. tnt Bd Fingl. 1982, 21, $O1; (1) Mulder, J. J. Co; Oosterhoff, L. J. Chem, Commun. 1970, 805, 307; (g) van der Hart, W. J: Mulder J.J. Cs Oosterholf, LJ. f-Aw. Chem, Soc. 1972, 94, 5724: (h) Goddard, W. A. IL, J. Am. Chem. Soc. 1972, 94, 795; (i) Dewar MJ. S: Kirschner, 8: Kollmar, H, W. J. Am, Chem. Soc. 1974, 96, 5240 and following papers; () Silver, D. M. J. Am. Chem, Soc, 1974, 96, 5959; (k) Epiatis, N. D. “Theory of Organic Reactions.” Springer-Verlag: Berlin, 1978; () Halevi, E. A. Angew. Chem. Int. Ed. Engl. 1976, 15, 503 general: (m) Pearson, R. G. "Symmetry Rules for Reactions,” Wiley: New York, 1076; (a) Fukui, K. “Theory of Orientation and Stereoselection,” Springer-Verlag: Berlin, 1975; (0) Houk, K. N. Ace. Chem. Res, 1975, 8, 361, (p) Gilchrist, T. Ly Storr, R. G. “Organic Reactions and Symmetry,” Cambridge Univer London, 1972; (q) Fleming, 1."Frontier Orbitals and Organic Chemical Reactions," Wiley: London, 1976, (1) Marchand, A. Pa Lehr, RE. Eds. “Perieyclic Reactions,” Vols. | and 2; Academic Press: New York, 1977; () Metiu, H.; Ross, J. Whitesides, G. M. Angew. Chem, tnt Ed. Engl. 1979, 18, 977: () Epiotis, Nii Shatk, Si Zander, W. in “Rearrangements in Ground and Excited States"; De Mayo, P., Ed.; Academie Press: New York, 1980, Vol. 2, p. 2 "See note a. *Hiuingen, R. Angew, Clem, Int. Ed. Engl. 1968, 7, 321. Hluisgen (note 5) designated componends according (o the number of atoms rather than number of electrons, but the notation used here is now universal 10.1 Definitions 841 (10.1) Sy ate [+ J=L | (10.2) f EO ~S FF Aan Ral = (104) 1 ‘The interesting feature of the cycloadditions is that the ease with which they take place depends on the number of electrons involved in the bonding changes. Whereas the 2 + 4 process (Equation 10.2), known as the Diels—Alder reaction, occurs readily with activation free energies of roughly 20-30 keal mol! (about 84 to 126 kJ mol '), and has been one of the cornerstones of organic chemistry for many years,’ the 2 + 2 and 4 + 4 additions are accomplished much less easily. One finds, however, that ifone of the reacting molecules is in an electronically excited state, the 2+ 2 and 4+ 4 processes occur more readily than the 2+ 4.8 The 1,3-dipolar additions (Equation 10.5), studied and co-workers,” also fit in the cycloaddition category. Th xtensively by Huisgen dipole, which can 244 [+ (10.5) —O), nittile oxide (—C=N—O: —C=N—O:) —typically has four a elec- trons in an orbital system delocalized over three centers. These reactions, like the Diels—Alder, are easily accomplished. (a) Kloetrel, M. CG. Org. Reactions 1948, 4,1; (b) Holmes, H. 1. Org: Reactions 1948, 4,60; (c) Bute, L. W.: Rytina, A. W. Org Reactions 1949, 5, 136; (@) Martin, J. G.s Hill, RK, Chem. Rev. 1961, 61, 37; (e) Huisgen, R.: Grashey, R.; Sauer, Jin The Chemistry of Alkenes”; Pati 8, Ed.; Wiley: London, 1964, p. 739; (Q) Wassermann, A. “Diels~Alder Reactions"; Elsevier Amsterdam, 1965; (@) Kw om, Rev. 1968, 68, A15. *See, for example, Dil 1969, 69, 845. Huisgen, R, Angew. Chem, Int, Ed, Engl, 1963, 2, 565. 842 THe THEORY oF PeRicyeitc REACTIONS Stereochemistry in Cycloaddition For a given z-clectron system there are two stereochemical alternatives, illustrated for the four-electron fragment (a diene) of a 2 +4 addition by Structures 2 and 3, in which the arrows show where the two new @ bonds are forming. In 2 both bonds A / form to the same face of the four-electron s whenever two bonds form on the same face of a unit entering a cycloaddition, that unit is said to enter suprafacially, abbreviated s. In 8 addition occurs on the upper face at one end of the diene and on the lower face at the other; in this instance, the dl ng antarafacially, a. In both Structures 2 and 8 the two-electron component (an alkene) is acting ina suprafa~ ay; in 2 we would say that the combination is (diene s) + (alkene s), and in 3 it is (diene @) + (alkene s). Further economy of notation is obtained by indicating only the number of electrons that enter from each unit. The diene is contributing four from its a system and so is designated 4y in 2 and 4a in 3, ‘The alkene is two electrons and so is designated 2s in each case. Hence the shorthand ne is enter cial y electron: contributin; notation summarizing number of electrons and stereochemistry is 25 + 4s for 2, and 25 + 4a for 3. Any of the combinations s + s,s + a, a + s,@ + ais conceivable for a cycloaddition of two components. Comparison of cis—trans stereochemistry of substituents in prod- uct to that in reactants establishes which occurred. In additions of relatively short chains the antarafacial interaction is difficult for the molecule to attain, but when systems with appropriate geometry are contrived, it is found that reactions in which one component acts in the antarafacial manner exhibit an inverted preference with respect to ring size: in the electronic ground state the 2 + 2 additions are now favora- ble and the 4+ 2 are not. Electrocyclic Reactions the second category is that of the electrocyclic reaction, illustrated in Equations 10.6— 10.8, characterized by opening or closing of a ring within a single molecule by conver sion of o bonds to 7 bonds or the reverse. Again, a notation with curved arrows (4) is helpful if used with appropriate caution. The primary feature of interest here is the stereochemistry of substituents of the o bond that is broken. These substituents lie 10.1 Definitions 843 (10.6) (10.7) (10.8) (10.9) above and below the ring plane, and as the bond breaks must move in one of the illustrated in Equation 10.10, Woodward and Hoffmann introduced the terms Ser Ei S28 (10.1 aes, ‘ » 6 % wi disrotatory conrotatory irections) and conrolatory (rotating in the same direc tion) to describe these alternatives." Again the preferred path depends on ring size: clectrocyclic ring opening and closing of a ground-state cyclobutene (Equation 10.7) usually follows the conrotatory route, whereas a ground-state cyclohexadiene (Equation 10.8) prefers the disrotatory. The preferences reverse in the photochemical excited-state reactions, It will be useful later to be able to apply the suprafacial—antarafacial notation to electrocyclic reactions, and to the other pericyclic reactions described below, as well to cycloadditions. Structure 5a represents schematically a7 bond interacting supra ially; the arrows indicate that interactions are occurring from the 7 bond shown, on the same face of the 7 bond at each end, to some other unspecified bond or bonds Structure 5b shows how this concept extends to a diene system. Structures 6a and 6b disrotatory (rotating in opposite d "Woodward, RB; Hoffmann, Rj. Am. Chem. Soc, 1965, 87, 395, 844 THe THrory or Pericyeiic REACTIONS HR Mg bal ging, eat show the 7 antarafacial interaction, in which the 7 system interacts at the two ends on opposite faces. If we represent a ¢ orbital in the analogous way as being made up of two hybrids, we can define @ suprafacial interactions by 7 and @ antarafacial interac tions by 8. The analogy between the m and @ systems may be made clear by rotating the g orbitals in 7 and 8 by 90° and expanding the small lobes to change hybridization from sp” to p. We can also define suprafacial and antarafacial interactions for a single orbital as illustrated in 9. 2 suprafacial o antarafacial 7 8 Leech © suprafacial co antarafacial With the help of the definitions illustrated in Structures 5-9, we can consider clectrocyclic reactions as a category of cycloaddition. For example, in the disrotatory ring opening depicted in Equation 10.11, a «bond consisting of orbitals | and 2 enters suprafacially into a new interaction; the other partner in the reaction is the system of two 7 bonds, orbitals 3, 4, 5, and 6, which also enter suprafacially. This process can therefore be thought of as a generalized cycloaddition of type os + 14s. Equations 10.12 and 10.13 show two equivalent ways of looking at a conrotatory electrocyclic reaction, 10.1 Definitions 845 30 (10.11) Y Qs —s sp - 8 » » —_— bg — (10.13) S ee b —— b (10.30) 28 oye [CEN 29 10.2 Pericyclic Reactions and Transition State Aromaticity 853 2s oe Be a (10.32) a € ry -¢ Ae - C : The next step in the analysis is to establish rules for assigning aos signs to the orbitals of interaction diagrams. Whenever a set of basis orbitals is chosen, the initial choice of sign is arbitrary. In Chapter 1 we chose basis function signs, as a matter of convenience, in such a way that bonding molecular orbitals would be the algebraic sum and antibonding orbitals the algebraic difference of the basis orbitals, as shown below for Hs: On. Satin Pig Po mae @ ®@ . . Pun ten vonding combination Pig + ley is 854 ‘Tue Tory oF Pertcyciic Reacrions We could just as well have made a different choic it Co) = ‘The higher-energy molecular orbital still comes out to be the one with a node; the nly difference is in its mathematical expression in terms of the basis. Given that the initial choice of basis orbital signs is arbitrary, we are free to adopt the following convention: Basis orbital signs of interaction diagrams are to be chosen in such a way that the interaction path connects only lobes of like sign so far as possi- ble. action diagram 31, Suppose we set the signs of the nt does not conform to the Take as an example the inte orbitals at random, say as shown in 32. This assig 3} sc convention stated above: there are two points, marked by asterisks, at which interac- tion paths connect lobes of unlike sign. A connection between lobes of unlike sign is called a phase inversion. Recalling that the interaction path is the fundamental aspect of the diagram, derived from the topology of the transition state under consideration, and thus cannot be changed but that the signs are arbitrary, we simply make a diffe ent choice of signs so as to eliminate the phase inversions (33). Now look at diagram f vy 22 oy § 8 10.2 Pericyclic Reactions and Transition State Aromaticity 855 28. There is no way of choosing the signs so as to avoid a phase inve:sion; the best we can do is 34, Other sign choices would only move the phase inversion to another point (34a) or create more phase inversions (4b). Cyclic interaction diagrams 33 and 34 belong to two fundamentally different classes: diagrams with no phase inversions, and diagrams with one phase inversion. as a3 8 8 yh Ty te bet 34e de It is easy to sce that there are only these two types of monocyclic interaction diagrams. If a particular lobe has > inversion on one side but not on the other, changing the sign of that lobe moves the phase inversion from one side to the other but does not change the number of phase inversions in the ring. (The change 34 — 34a is an example.) If a lobe has phase inversions on both sides, changing the sign of that lobe eliminates both phase inversions. (The change 34b — 34 is an exam- ple.) Therefore if an initial arbitrary assignment of orbital signs in a cyclic interaction diagram produces phase inversions, they can be moved around by changing signs until they meet on either side of a single lobe and can then be eliminated in pairs. (The series $4¢ > 34d — 34e illustrates this process.) All rings with an even number of phase inversions are thus equivalent to rings with no phase inversions, whereas all rings with an odd number are equivalent to rings with one. Rings with zero (or an even number of) phase inversions are said to be of the Hiickel type; rings with one (or an odd number of) phase inversion are said to be of the anti-Hiickel or Mdbius type. Aromatic and Antiaromatic Transition States Consider the ground-state benzene molecule in ter orbitals are arrayed on a ring, each interacting with its neighbor (35). The transition ns of an interaction diagram. Six p DRY OF PERIGYCLIG REACTIONS _¢ i 35 state interaction diagram of the #5 + ads cycloaddition (33) is identical. The aron ticity theory of perieyclic reactions holds that the 72s + ms transition state will enjoy a stabilization of the same kind as does benzene ground state. (Stabilization of the same kind, but not of the same magnitude, because distortions are required to make the interaction diagram of the pericyc tion state conform to a regular ring pat- tern.) The 72s + ms process with (4 + 2)m electrons (1 = 1) hi tion state and is thermally allowed. This analogy between benzene and the Diels— Alder transition state was first pointed out by Ev 1939;"" its applicability to 3 wide variety of reactions, however, was not fully appreciated until after the pioneering work of Woodward and Hoffmann, which we describe in Section 10.4. The cyclobutadiene molecule has the interaction diagram 36. The transition state for a 72s + 72s cycloaddition is equivalent; this 4 z-electron system (n= 1) omatic, the transition state is destabilized, and the reaction is thermally fd ic trans an aromatic transi-~ ans 36 Note that the m2s + mds and ms + 72s examples both have interaction diagi of the Hiickel type, without phase inversions. Our conclusion is that pericyclic reac- tions taking place through transition states of Hiickel type when thermally induced from the ground state are allowed for 4n + 2 electrons and forbidden for 4n elec trons, We must now consider how the situation changes for an ar in 1959 showed that the aromatic character of a ring with a given number of 7 elec- trons would change if the conjugated system included d orbitals.'' He considered systems with an equal number of d and p orbitals, for example, rings composed of alternating phosphorus and nitrogen atoms; the point of importance to the present discussion can be illustrated more effectively with just one d orbital in interaction ums 'Sgvans, M. G. Trans, Paradan Soe. 1999, 35, 824 Craig, D. PJ. Chem. Soe. 1959, 907 10.2 Pericyclic Reactions and Transition State Aromaticity 857 diagram 37. Given the interaction pattern specified, there is a single-phase inversion. Heilbronner pointed out that the same situation occurs in a singly twisted ring (38 or 39) containing only p orbitals.'® The topology is that of the Mobius strip, obtained by Sas CAFR FEET I? 37 38 39 joining the ends of a su'ip containing a single twist. For the Mobius systems, if one ignores the loss of stabilization caused by the twisting, the Hitckel 4” + 2 rule is versed: The twisted rings are aromatic for 4n and antiaromatic for dn + 2 electrons. It is the presence of the single phase inversion that causes this reversal. In Section 10.3 we shall present simple arguments, due to Dewar and to Zimmerman, that justify this conclusion, The as + 72a cycloaddition illustrated in Equation 10.30 has a single phase version i ction diagram (84). This transition state is therefore of the anti- Hiickel type. Because there are dn clectrons the reaction has an aromatic transition nd is thermally allowed We may summarize the procedure for predicting the allowed or forbidden natu of pericyclic reactions by the aromaticity method as follows: re- 1. Identify the cyclic path of interacting orbitals in the transition state and con- struct the interaction diagram. 2. Assign algebraic signs to the orbit inversions. 3. Classify the type of ring: als So as to minimize the nui nber of phase zero phase inversions: Hiickel ‘one phase inversion: anti-Hiickel 4. Predict whether the reaction is thermally allowed or forbidden: Hiickel rings: 4m + 2 electrons; thermally allowed 4n electrons, thermally forbidden kel rings: 4n electrons, thermally allowed 4n + 2 electrons, thermally forbidden anti-Hii Having found a set of pericyelic selection rules, we must add a few cautionary remarks. First, the rules apply to ground-state, thermal reactions only. The Wood- ward—Hoffmann procedures, which we shall examine in Section 10.5, and Zimmer- man’s method predict reversal of the rules for photochemical reactions, which occur 'Heilbronner, E. Tetrahedron Lett, 1964, 1923, 858 Tue THeory oF Pericyciic REACTIONS: from excited-state molecules. Second, the rules are not absolute. The analysis consid- cred only the electronic interactions of the pericyclic system and did that in the con- text of an interaction diagram that in some instances requires distortion to bring it into conformity to the ideal, regular planar polygon to which aromaticity theory ap- plies. Steric effects and the energy costs of nonideal orbital overlap are not taken into account nor are entropy changes. ‘These simplifications mean that further considera- tions, beyond the pericyclic theory, will sometimes need to be applied. As we shall see in Chapter I1, the predictions of the theory are in general well substantiated by experiment, but there are many circumstances under which “allowed” reactions have high activation energies for steric reasons, or “forbidden” reactions occur more easily than expected because of relief of ring strain or some similar factor. When these extra effects can be taken into account, better tests of the predictions of pericyclic theory are possible. 10.3 AROMATICITY AND ANTIAROMATICITY IN HUCKEL AND ANTI-HUCKEL RINGS Zimmerman approach to the stabilization of pericyclic transition states, the basic criterion for deciding whether a process will be allowed or forbidden is whether the transition state is stabilized or destabilized, respectively, compared with an analogous open-chain system.'® As we have noted above, this criterion is the same as that for aromatic or antiaromatic character of a ground-state system. In the appen- dix to Chapter 1, we showed how Dewar’s perturbation molecular orbital (PMO) the- ory accounts for icity by comparing the energy changes that take place when a chain with an odd number of p orbitals is linked either at one end or at both ends to a single carbon p orbital to make either an even-membered chain or a ring.'? Figure ALG (p. 119) showed how the energy stabilization of these processes is primarily de- pendent on the nonbonding molecular orbitals (NBMO), which are the frontier orbit- als for the fragments, We review that argument in Figure 10.1. Each union made between the fragments entails an orbital energy lowering of AE = aoiaoaB (10. In the Dew: ) where the do, are the coefficients of the nonbonding frontier molecular orbitals of the two fragments at the point where the junction is made. Since the coefficient of the isolated p orbital is unity and the coefficient at the end of the allyl chain is 1/V2, the formation of butadiene leads to energy lowering AE [Figure 10.1(a)] of B/V2, and stabilization (for two electrons) by twice this amount, or V2. When the union takes place at both ends of the allyl system, however, the two energy changes are of opposite sign because of the node in the ally! NBMO. Hence no stabilization occurs and the cyclobutadiene system is destabilized relative to the open-chain analog butadiene. The same analysis applied to formation of a six-membered conjugated ring (benzene). will show net stabilization relative to the open-chain hexatriene (Problem 11 in Chapter 1) (a) Dewar, M. J. S. Tetrahedron Suppl1966, 8, 75; (b) Dewar, M. J. 8. Angew Chem, In. Ei Engl 1971, 10, 761. (a) Dewar, M. J. §. "The Molecular Orbital Theory of Organic Chemistry"; MeGraw-Hill: New York, 1969, p. 217: () Dewar, M. J. 8 Dougherty, R. G. “The PMO Theory of Organic Chemistry"; Plenum: New York, 1975, p. 59 10.3 Aromaticity and Antiaromaticity in Hitckel and Anti-Hackel Rings 859 I with one end of Figure 10.1 (a) Perturbation analysis of the union of a single carbon p or interaction of the p an allyl a system. The principal energy change arises {rom orbital with the allyl NBMO, which results in a stabilization (fer two electrons) of 2AE = V2 B for union at one end to form butadiene. (b) When ui at both ends to form cyclobutadiene, the opposite signs of the allyl NBMO coefficients at the two ends lead to AE = 0 and no net stabilization. Cyclobutadiene is stabilized less than its open-chain analog butadiene and is antiaromatic 860 THe THrorY oF Pericyetic REACTIONS ‘The utility of the Dewar PMO method for our present purposes lies in its ability to clarify the situation for anti-Hiickel rings. If there is a phase inversion in the ring of interest, we are free to choose basis function signs so as to put that phase inversion at one of the points of union. Figure 10.2 shows such a situation; our earlier analysis of interaction diagrams should make it clear that this picture is the appropriate one for the transition state of a conrotatory cyclobutene ring opening. The phase inversion i the interaction diagram translates to a reversal in the sign of the Hiickel inte parameter B; the stabilization for the interaction shown in Figure 10.2 is therefore 1 i E= = V9 35) AE = (8) ~ a (-P) = V28 (10.35) an energy stabilization of 2V2B, which is V2 greater than that ndard, Hence anti-Hiickel cyclobutadiene is aromatic; the same -Huickel benzene will show that it is antiaromatic. or, for two electrons, of the open-chain st analysis applied to Zimmerman has presented a similar XN Figure 10.2 Perturbation analysis of the formation of anti-Hiickel cyclobutadiene. At the top, the basis orbitals are shown, with an interaction diagram imposed that generates a phase inversion. The phase inversion results in the reversal of the sign of B for the union at that point; the result is stabilization (for two electrons) of 2V2B, which is greater than for the open-chain model in Figure 10.1 (a). Anti-Hiickel cyclobutadi- ene is thus concluded to be aromatic. 10.3 Aromaticity and Antiaromaticity in Hiickel and Anti-Hiickel Rings 861 ‘Transition state Starting state Product stat two x bonds of two o bonds of HaC—cHy 2H,C=CHy Enersy 7+ Hp Figure 10.3 The Zimmerman analysis of the #2s + 72s cycloaddition. The diagram traces the energies of the reacting orbitals from starting state (left) through transition state (cemer) to product (right). At the left, two bonding ethylene a orbitals are occu- pied by pairs of electrons. In the transition state there is only ore bonding orbital; one pair of electrons is forced to climb an energy hill to the nonbonding level. Ln the product both electron pairs are again in bonding orbit js time or orbitals of cyclobutane. 15,0 between the pericyclic transition state energy levels and the Hiickel 7 energy levels.'* If one starts with two ground-state ethylenes, as shown at the leftin Figure 10.3, there will be two pairs of electrons in the two bonding 7 molecular orbitals. The transition state for their union has the eyelic interaction equivalent to cyclobutadiene; it there- () Zimmerman, HE, Acc. Clem Ree. 1971, 4, 272: nd, A. P.. Lehr, R. E. Eds; Acacemic Press: New York, liatomaticity idea to pericyelic reactions. See note 12 SG) Zimmerman, H. E. J. Am. Chem. Soc. 1966, 88, 1564, 156 (6) Zimmerman, H. E. in "Perieyelic Reactions”; Vol. I: March 1977, p. 53. Zimmerman was apparently the first to apply th 862 ‘Tie Trirory oF Pericyeiie Reactions fore has one bonding MO and two degenerate nonbonding MO’s. Figure 10.3 waces the energies of the relevant orbitals as the two ethylene molecules come together. One pair of electrons can remain in a bonding MO, but the other is forced to climb an energy hill to the nonbonding MO of the transition state before coming back down to a bonding & orbital in the product cyclobutane. The antiaromatic nature of the wansi tion state thus imposes a substantial energy barrier, and the model explains the high activation cnergy of the reaction. This argument is extended by analogy to include other perieyelic transi As Figure A1.5 (p. 115) in the Appendix to Chapter I showed, the regular pattern of energy levels of cyclic conjugated systems has degenerate, nonbonding energy levels in pairs for each of the dn systems. ‘Thus other 41 pericyclic transition states will be destabilized in the same way as the 2+ 2 cycloaddition, Zimmerman has demon- strated that a circle mnemonic device similar to that of Figure A1.5 applies to anti- Hiickel rings. Figure 10.4 shows this scheme, in which polygons are inscribed in circles of radius 28 with a side instead of a corner downward. The energy levels are again found by horizontal projection from the vertices; now nonbonding degenerate levels, and hence antiaromatic transition states, occur for 4a 2 rings. The Zimmei analysis is thus equivalent to the one given above based on Dewar’s methods Several other investigators have also used aromaticity theory to analyze pericyclic reactions." jon states. 10.4 FRONTIER ORBITAL INTERACTIONS IN PERICYCLIC REACTIONS A second approach to pericyclic reactions uses perturbation theory and symmetry to analyze interactions of the frontier orbitals (highest occupicd molecular orbital, HOMO, and lowest unoccupied molecular orbital, LUMO)."° This is the method Woodward and Hoffmann used in their first paper on pericyclic selection rules;*! it has subsequently been developed by Fukui and others.” Symmetry in Pericyclic Reactions In Section 1,7 we discussed the formation of symmetry adapted molecular orbitals and showed how they are classified as symmetric (S) or antisymmetric (A) with respect to symmetry operations of the molecule.2* We also illustrated the use of symmetry in assessing interactions among orbitals: only orbitals of the same symmetry can interact Figures 10.5 and 10.6 show examples that review the concepts of orbital symme- try classification. Figure 10.5(a) shows the a bonding orbital of ethylene together with Gq) Goldstein, M. Jz Hoffmann, R.J- Am. Chem. Soe. 1971, 93, 6193; (b) Salem, L- J. Am. Chem. Soc. 1968, 90, 54 (0) Mulder, J.J. C Oosterhoff, LJ. Chem, Commun, 1970, 305, 307; (d) van der Hat, W. J Mulder, J-J- G3 Oosterhoff, L. J. J. Am. Chem. Soe. 1972, 94, 5724; (c) Herndon, W. G, Chem. Rev, 1972, 72, 157. The idk symmetry properties of orbitals might be important in determining the course of certain reactions appar- ly originated with L. J. Oosterhoft, quoted by Havinga, E.z Schlatmann, J. L. M.A Tetraedvan 1961, 16, 146. Woodward, R. B.: Hoffmann, R. J. Am. Gem. Soc. 1968, 87, 395, 2 (a) Fukui K, “Theory of Orientation and Stereoselection”: Springer-Verlag: Berlin, 1975; (b) Houk, K.N. Ae. Cheam. Re 1975, 8, 361; () Fukui, K. Angew. Chem. Int. Ed. Engl, 1982, 21, 801; () Houk, K. N. in “Pericyclic Reactions"; Vol. I Marchand, A. P, Ler, R. E., Eds; Academic Press: New York, 1977, p. 181 When there are axes of symmetry of threefold or higher, there will be orbitals that tr We shall not have eecasion to use these higher symmetries in our discussion. nsform in more complicated ways, 863 10.4 Frontier Orbital Interactions in Pericyclic Reactions “Duy ‘sso4q onmopeoy Jo iad dq pordepy ¥9 -d {2161 SOA MON: fSPY AW AUST “Ad “Vy “pueyauEyy “104 ‘suonmay mpaouag woLy Murad jason “tite jo sorSiouio areas uonsuien 405 sfopow se pasn age sjaray B1aua ay “ZA =P GBA —? ZA +2 ‘ABA +” soruaua sey ‘aduiexa 105 “Bu smoy SY], ‘sjaaa] ABiaua 2’ ayp saat A[peruor10y uondoload 4 Sopa1) Ur UMop apis paqurosut are suo’ dave +. sco +0 “y seert + as9L10 + — — stro +0— —-,| yet #s910-»— —.. pee pez —-»>— —-- dsre't »— 8 (es zev1-»— —- scams devi» §4+2——_. siF0—28— —--4 SrIvl +9 a-» — g—0 864 Tur Troy oF Penicyciic REACTIONS Oo) Oy Figure 10.5 (a) The ethylene # bonding orbital is symmetric (S) with respect to reflection in the mirror plane o, and antisymmetric (A) with respect to reflection in the mirror plane o. (b) The ethylene 7* antibonding orbital is antisymmetric (A) with respect to reflection in both o and o' two of the symmetry clements of the molecule, the mirror planes o and o!. This orbital is symmetric (S) with respect to reflection in plane @, but antisymmetric (A) with respect to reflection in the plane o’. The 7* orbital [Figure 10.5(b)] is of a differ- ent symmetry type, antisymmetric (A) with respect to both reflections. Figure 10.6 shows the 7 orbitals of butadiene, which are symmetry adapted with respect to the two mirror planes and the Cy axis that are symmetry elements of the molecule in the s-cis conformation. To avoid confusion we must emphasize that Figure 10.6 depicts molecular orbitals. Although the drawings should indicate extended lobes (40 and 41 for 7 and 7), the difficulty of making such diagrams, especially in the more complex structures, dictates a simplified notation in which one merely indicates the relative signs with which the basis functions are combined in each MO, Molecular 40 41 orbitals shown in the manner of Figure 10.6 look very similar to the diagrams we used in Section 10.2 to show sets of basis orbitals. In the remainder of this chapter we shall be discussing molecular orbitals and the diagrams will show molecular orbitals unless otherwise specified. Figure 10.6 shows that the 7 orbitals of butadiene all have the same symmetry classification, A, with respect to the mirror plane o'. Classification with respect to 0! therefore does not contribute to distinguishing among the four orbitals. The purpose 10.4 Frontier Orbital Interactions in Pericydic Reactions 865 Symmetry elements Molecular orbits Symmetry C0" i V5b9 —_ a LOR — asa Enersy n ‘ ‘ _— ‘SAA Figure 10.6 The 7 molecular orbitals of butadiene. At the top are shown the symmetry ele ments of the molecule, wo mirror planes and a Cz axis. Below are the four molecu- Jar orbitals in an cnergy-level diagram, with their symmetry behavior under each of the symmetry operations listed at the right. Note that the diagrams are of molec- ular orbitals shown as composites of the constituent basis orbitals, each with its appropriate relative phase. so that interactions of symmetry classification is to separate the orbitals into categories can be assessed; symmetry elements that do not assist in this classification add no useful information and are therefore usually omitted. Such omissions are made in some of the examples we consider below; the extra symmetry elements can always be included, but they will contribute nothing new to the argument. We also need another idea, already introduced in Section 1.9, that in some circumstances itis appropriate to 866 ORY OF PERICYCLIG REACTIONS use a symmetry element that is not strictly, but rather only approximately, a correct symmetry element of the molecule. The reason we can do this is that in pericyclic reactions we focus on those orbitals in the molecule that are actually involved in the bonding changes of interest, the reacting orbitals. The system of reacting orbitals may have an intrinsic symmetry that will be appropriate to use, but there may be a substitu ent in the molecule that, technically speaking, destroys this symmetry. If the substitu ent does not interact strongly with any of the reacting orbitals, the situation should be the same as if the substituent were not there, and the higher symmetry of the reacting orbitals themselves will be the proper one to use. Let us suppose, for example, that we wish to know about the interaction of the 7 orbitals of butadiene with orbitals of some other molecule. If the butadiene is unsub- stituted (Figure 10.6), the molecular symmetry and the symmetry needed for the orbital model are the same. If we now consider 1,3-pentadiene, Figure 10.7, we find that the introduction of the methyl group has removed two of the symmetry clements We may nevertheless argue that the methyl group will have only a secondary effect on Molecular orbitals Energy £5 25 Figure 10.7 The = orbitals for 1,3-pentadiene. Now the correct molecular symmetry consists of only the plane a’, but the z orbitals are still constructed according to their approxi- mate local symmetry, which remains as shown in Figure 10.6 10.5 Correlation Diagrams 867 the 7 system, the main features of which should be approximately the same as in butadiene itself. For purposes of qualitative arguments, we can therefore ignore the methyl substitution and take the symmetry of the a system as the applicable try. The proper symmetry to use in this instance, then, is the same as that of the unsubstituted butadiene. In using approximate symmetries we are doing much the same thing we did in constructing interaction diagrams. We are identifying the intrin- sic symmetry properties of the set of reacting orbitals Urat constitutes the energy- determining part of the interacting system. We have discussed in Section 1.8 the application of perturbation theory to processes in which two molecules come together. We saw there that the most impor- tant interactions will be between the HOMO of one molecule and the LUMO of the other. This method can serve as a useful guide in deciding whether there will be yclic reaction begins to occur. Let us consider ethylene and shown in 42. We wish to consider a stabilization as a per butadiene approaching each other in the manner ed the interactions of the HOMO of ethylene with the LUMO of butadiene, and of the LUMO of ethylene with the HOMO of butadiene. In order to do this, we identify first the symmetry of the aggregate of the two molecules arranged as in 42; there is one symmetry element, a vertical mirror plane that cuts across the ethylene double bond and the butadiene single bond. In Figure 10.8 we place at the left the 7 orbitals of butadiene, and at the right the 7 orbitals of ethylene. The symmetry classification of each orbital is with respect to the mirror plane shown. (Note that the other symmetry elements that would be applicable to butadiene alone or to ethyleae alone are not correct symmetry elements for the combination 42 and are not included.) The interac- tion of each HOMO with the other LUMO is permitted by symmetry; as we have seen in Section 1.8, this is just the situation that leads to stabilization, and we may expect this cycloaddition to he facile, as indeed it is If we now look at the cycloaddition of two butadiene molecuks to cach other (Figure 10.9), we find that because of the symmetry mismatch between the HOMO of one molecule and the LUMO of the other, there can be no stabilization. ‘The only illed levels, which, as we have seen in Section 1.6, is destabiliz~ ss should not occur readily, a conclusion that is again in agree- interaction is between ng. Hence this proc ment with experiment 10.5 CORRELATION DIAGRAMS Although their initial paper looked at pericyclic reactions in terms of frontier orbitals, Woodward and Hoffmann’s subsequent d have for the most part used or- 868 ‘THe THrory oF PericycLic REACTIONS Symmetry elements Butadiene Ethylene Molecular orbitals ‘Symmetry ‘Symmetry Molecular orbitals ie LUMO Energy A HOMO ts EF 35 Figure 10.8 HOMO-LUMO interactions in the approach of ethylene to butadiene. The sym- metry (9) is shown at top center. At left are the butadiene MO's, classified ac- cording to their symmetry with respect to 0; at right are the ethylene MO's, also classified according to ¢. The HOMO of each molecule can interact with the LUMO of the other, and a stabilization occurs as they approach one another. 10.5 Correlation Diagrams 869 Symmetry elements Butadiene 2 Butadiene 1 Molecular orbitals Symmetry Molecular orbitals A i Energy Figure 10.9 The approach of two butadiene molecules. The symmetries do not permit HOMO-LUMO interaction; the interaction between filled levels, permitted by the symmetry, gives no stabilization. 870. ‘The Tirory oF Prnicyciie REACTIONS bital correlation." In the orbital correlation method one first identifies in the starting molecules and in the products those molecular orbitals (the reacting orbitals) that correspond to bonds being formed or broken during the pericyclic process. ‘To make the connection with the aromaticity analysis in Section 10.2, we note that these molec- ular orbitals are just the ones constructed from the basis atomic orbitals that make up the interaction diagram for the reaction. It is, in fact, through the interaction diagram that one may establish the equivalency of the orbital correlation and the aromaticity approaches to pericyclie theor Having identified the reacting orbitals, we next find those symmetry elem the system that are maintained throughout the reaction, The reacting orbitals then classified according to their symmetries with respect to these elements and are traced (correlated) through the reaction from starting materials to products, always conserving their symmetry properties. principles with seve nts of re We shall illustrate thes electrocyclic interconversion of butadiene and cyclobutene, looked at in the (non- . Consider first the exampl Il be recalled that this ing direction (Equation 10.36). It w CI process can occur in either of two ways: disrotatory (Equation 10.37) or conrotatory (Equation 10.38) and that the thermal reaction exhibits a marked preference for the conrotatory path. spontaneous) (10.38) The reacting orbital system needed to include all the bonds being formed or broken is made up in the reactant from a basis consisting of a p orbital on each of the four carbon atoms, and in the product of p orbitals on each of the wo t-bonded carbons and a hybrid orbital, approximately sp", on each of the two carbons linked by the newly formed o bond. Figure 10.10 illustrates the reacting molecular orbitals. Consider now the changes that will occur when a conrotatory ring closure takes, place. Figure 10.11 shows that as soon as the end carbons of the diene begin their (a) Hoffmann, Rs Woodward, R. B. Aee. Cem, Res. 1968, 1, 17; (b) Woodward, R. Bu Hof ‘of Orbital Symmetry”; Verlag Chemie: Weinhein/Bergstr., Germany, and Acaderni orbital correlation theory is in some respects more complex than the Dewar-Zimmerman aromaticity theory, it allows better insight into the energy changes of individual orbitals during the pericyclic reaction and gives a better understanding of excited-state reactions ann, R. “The Conservation Press: New York, 1970, Although the 10.5 Correlation Diagrams 871 a Reactant “ Product oho. a 9388 . Figure 10.10 The orbitals of reactants and products in electrocyelic ring closure of butadiene. rotation, the symmetry of the reacting orbitals changes. The two mirror planes are no longer present, and only the Cz axis remains. We now wish to follow the molecular orbitals through the transformation using the principle that a molecular orbital main- lains its symmetry during a bonding reorganization. Figure 10.12 reproduces the important molecular orbitals and classifies them 872 THe THeory oF PericyetiG React ONS =) 5) Figure 10.11 Symmetry at an intermediate stage of the conrotatory closure. Only the G2 axis remains as a symmetry element. according to their symmetry with respect to the Cy axis, the element that defines the symmetry during the conrotatory process.®” Orbital 7, antisymmetric under C change continuously into an orbital of the product in such a way as to remain at all stages antisymmetric under the Cy operation. ‘The symmetry conservation principle allows us to reconstruct qualitatively how the orbital 7, will change. Since it starts out antisymmetric under C2, it remains so; it can do this only if it ends up as an antisym- metric orbital of the product, say a, As the two end carbons rotate, the contribution of the p orbitals on those end carbons must decrease, finally to disappear altogether. In am, symmetry type 5, the contribution of the two central p orbitals will decrease, leav- ing only the end two, which will have rotated onto each other to yield the product & orbital. The changes in the antibonding orbitals may be visualized in a similar way. “The lines joining reactant and product orbitals in Figure 10.12 are referred to as covvelation lines, and the entire diagram is an orbital correlation diagram. It will be noted that since there are two orbitals of each symmetry type on each side, there is an alternative way the correlation might have been made, namely, 7 too, my 10 7, ai to m, wi toa. This alternative is eliminated by the noncrossing rule: Orbilals of the same symmetry do not cross Because the noncrossing rule is of fundamental importance to the construction of correlation diagrams, we digress a moment to give a justification for it, We have already seen in Section 1.6 that orbitals of the same symmetry interact in such a way as to push the lower-energy one still lower and the higher-energy one still higher, and that the interaction is stronger the closer the two are in energy. If two orbitals of the mmeury were to approach each other in energy, the interactions would there- fore tend to keep them apart and prevent the lines from crossing. The noncrossing In Figure 10.1 product lev Ary of the reaction and in subsequer Tepresented only eneryy-level diagrams we shall construct, the relative energies of reactant matically. One cannot deduce from diagrams of this type the overall thermochs 10.5 Correlation Diagrams 873 Symmetry elements QE a a faNS Q Q A—— a oy 5. “ ) a. A Figure 10.12 Classification of the reacting molecular orbitals of butadiene and cyclobutene for the conrotatory process. Symmetry classifications are with respect to the Co axis, $ indicating symmetric and A antisymmetric orbitals. The correlation lines are ob- tained by connecting orbitals of the same symmetry. 874 Tue THrory oF Pericyciic REACTIONS rule does not apply to orbitals of different symmetries, because they do not interact. Hence correlation lines representing orbitals of different symmetry may cro: Although the noncrossing rule may in many instances be relied upon to deter- mine the correlation pattern where alternatives exist, it is not infallible. In order to avoid difficulties in constructing correlation diagrams, Woodward and Hoffmann cite three precautions that should be observed.?° L. Processes that are inherently independent must be considered separately even if they occur in the same molecule. 2. Each reacting system must be reduced, by removing substituents and distor- tions, to its highest inherent symmetry. 3, Symmetry elements used in the analysis must bisect bonds made or broken in the reaction. Having obtained the orbital correlation diagram for the butadiene closure, we can now see that the electron pairs that start out in bonding orbitals of the reactant are uansferred into bonding orbitals of the product without encountering any symmetry imposed energy barrier. The proce id to be symmeiry allowed. This principle lies at the heart of the theory of orbital symmetry control of reaction path: A reaction is allowed in the ground state (thermally allowed) only when all reactant bonding electron pairs are transferred without symmetry imposed barrier into bonding orbitals of the product.*” The Disrotatory Electrocyclic Reaction We turn now to the disrotatory closure. The orbitals in reactants and products are the same as before, but this time it is the mirror plane o (Figure 10.13) that is maintained throughout. Figure 10,14 shows that the symmetry dictates correlation of a bonding to an antibonding orbital; the electron pair in orbital 7, if it remained in its original orbital, would end up in the high-energy orbital 7 of the product, a process that would require a large energy input. A similar situation would arise in the reverse reaction starting from ground-state cyclobutene, levels o and 7 occupied. In this case, then, the symmetry imposes a sizable barrier, and the thermal reaction is said to be symmetry forbidden Itis well to recall at this point that the terms allowed and forbidden in this context are not absolute. First, a symmetry allowed reaction will not necessarily take place with low activation energy. Other requirements, such as ease of approach, favorable over- lap of orbitals, and steric interactions may well combine to impose a substantial activa- tion barrier where none is predicted on the basis of symmetry alone. Moreover, favor- able arrangement of orbital energy levels may permit a forbidden reaction to occur more readily than might have been expected. Second, the symmetry rules are based on the assumption that the reactions are concerted. ‘There are always available © Woodward, R. B.; Hoffmann, 8. “The Conservation of Orbital Symmetry"; Verlag Chemie: Weinheim/Bergstr., Ger- many, and Academie Press: New York, 1970, p. 31 The energy changes of the orbitals in situations like uhat depicted in Figure 10. secondary interactions of the pericyclic orbitals with orbitals of substituent groups. For example, in conrotatory opening of cyclobutenes (right to left in Figure 10.12), if there are substituents with aunshared pairs at postions b in the cyclobutene, theres a stereoelectronic factor disfavoring rotation of the substituents inward (clockwise in this example) tat arises from interaction of the unshared pairs with the eyclobuitene « orbital. This effect is in addition to steric factors, which also disfavor inward rotation. See Rondan, N. G.; Houk, K. N. J. Am. Chem. Soc. 1985, 107, 2099. ase modified in their details by 10.5 Correlation Diagrams 875 Figure 10.13 Symmetry at an intermediate stage of the disrotatory closure. Only the mi nains as a symmetry element g through intermediates; the orbital symmetry predic- tions are vali aly if all the bonding changes occur together. It is instructive now to turn to the correlation diagrams in Figures 10.15 and 10.16 for conrotatory and disrotatory closure of hexatriene, a six 7-clectron system, The disrotatory mode is now thermally allowed, the conrotatory forbidden. If correla~ tion diagrams for larger systems are constructed, it will be found that with each add tion of two carbons and an electron pair the predicted selectivity will reverse. Photochemical Reactions ‘The correlation for a thermally forbidden reaction shows a bonding orbit y the highest occupied) correlating to an antibonding one (usually the lowest unoceu- pied). If absorption of a quantum of light of a suitable frequency raises the reactant to its first electronic excited configuration, the orbital occupancy will be as indicated in Figure 10.17. Now, during the reaction, as one electron goes up in energy, another comes down; correlation (assuming no switching of orbitals by the electrons) is directly toa product configuration of comparable energy to the starting configuration and the process should be allowed. Inspection of Figure 10.12 or 10.16 indicates, moreover, that an electronic ground-state allowed process should be forbidden photochemically because the lowest-energy excited configuration of reactants correlates to a more highly excited configuration of products. The orbital symmetry rules therefore pre- dict reversal of the allowed or forbidden nature of a given process when carried out photochemically. A word of caution is necessary here. Chemistry of electronic excited ‘Symmetry elements, Reactant Product ee aS e Ne Molecular, ‘Symmetry Symmetry: Molecular ‘stale : " ie a Ne’ : ate 5 : — r Energy a ae ne 5 sits Figure 10.14 Symmetry classification and correlation of orbitals for the disrotatory closure of butadiene. Closure with electron pairs remaining in their original levels would lead to the excited configuration indicated by the orbital occupancy on the right. Reactant ae Molecular orbitals aul Enengy n 10.5 Correlation Diagrams 877 ‘Symmetry elements Product tis Aa aQy, Cs é , Symmetry Symmetry Notecter a o real A o —" ° BEES i : ee AEN « 7 : 3o8$- ceed Figure 10.15 Correlation diagram for the thermally forbidden conrotatory closure of hexatriene. » REACTIONS ‘Symmetry elements. . Product x RS *otatr wat Symmetry, ‘Syanziry ; Enensy 10.5 Correlation Diagrams 879 Reactant Product Molecular Molecular orbitals orbitals Antibonding =. — —-— sn Bonding. ——4}—— $$$. ——#+— Bonding ly forbidden Figure 10.17 Orbital correlation diagram for a photochemically allowed, therm: pericyclic reaction, Absorption of a quantum of light raises on ron of the starting material from the highest occupied to the lowest unoccupied orbital. The system can pass easily to product because the energy cost of raising one electron is compensated by lowering the other. (Here we are approximating the lowest- energy excited state of reactant and product by the electron configurations shown. See Section 10.6 for further discussion of this point.) ood than that of ground states. ‘The simple correlation diagrams we have been using are inadequate for describing excited states in detail, and the predictions are less reliable than for ground-state reactions.** Chem- istry of excited states is discussed further in Section 10.6 and in Chapter 12. well under states is complex and le: Dougherty, RG. J. Am. Clem, Soc, 1971, 93, TIST has argued that systems whose ground states are aromatic have ic excited states and vice ¥ ie the universal criterion for allowed perieyelie reactions, both and excited state, is that the t ‘of our present knowledge of excited states nevertheless indicates that the more restricied statement given here is (0 be preferred. nl that there isition state be aromatic. ‘The uncert igure 10.16 (On lefi-hand page) Correlation diagram for the thermally allowed disrotatory elo- sure of hexatriene. The forms of the molecular orbitals indica'e correlations ac- cording to the dashed lines. The dashed-line correlations, however, would cause crossing of orbitals of the same symmetry. As two orbitals of the same symmetry approach each other in energy, they mix and interact strong\y. In this mixing process cach takes on character of the other. The orbital initial y at lower e: remains at lower energy while changing its form; similarly, the orbital initially higher energy remains higher. In this way the intended crossing is avoided. See ection 10.6 for further discussion of intended crossings 880 THe THEORY oF PERIcycLic REACTIONS Figure 10.18 Symmetry elements for the suprafacial approach of two ethylene molecules. Cycloaddition In Section10.1 we pointed out that a compound may enter into a cycloaddition in either of two ways, suprafacial or antarafacial. First we shall construct correlation diagrams for the simple 72s + 2s and 72s + m4s cycloadditions.”? Figure 10.18 illustrates the symmetry appropriate to the all-suprafacial approach of wo ethylenes. We have followed Rule $ of p. 874 in singling out for attention the symmetry clements that bisect reacting bonds. We have not included others, such as a Cy axis passing midway between the two molecules perpendicular to the page, that do not. (These symmetry elements not bisecting reacting bonds could be included with- ‘out affecting the outcome of the analysis. There is danger of error only when such elements are used exclusively.) The atomic orbital basis consists of a p function on cach of the four carbon atoms; Figure 10.19 illustrates the derived symmetry adapted molecular orbitals. As a result of the reflection plane that carries one ethylene into the other, these molecular orbitals are delocalized over both molecules. Figure 10.19 also shows the orbitals of the product; we consider only the two C—C o bonds formed and Hoffmann, Ri Woodward, R. B. J. Am. Chem. Soc. 1965, 87, 2046; (b) Woodward, R. B.; Hoffmann, R. “The Conservation of Orbital Symmetry"; Veriag Chemie: Wei Germany, and Academic Press: New York, 1970. Figure 10.19 Orbitals and orbital correlations for the ground-state 72s + 2s cycloaddition. In this example, at the left side of the diagram, reflection in the mirror plane 0! transforms the p orbitals of one ethylene molecule into those of the other; hence each molecular orbital of the combined system has a contribution from cach of the four p orbitals. The relative energies of these orbitals are shown after a small amount of interaction has begun; when the molecules are very far apart, SS and SA would be degenerate, as would AS and AA. 881 Diaj relation 10.5 Cor (Xn cee ae oe 882. THe THrory oF Pericyciic Reactions: ignore the other two, which were present from the beginning and did not undergo any change. When the two ethylene molecules are far apart, orbitals 7; and 72 are essentially the bonding z orbitals and will therefore each be occupied by an electron pair if both ethylenes are in their ground states. As the molecules approach, the intermolecu interaction of one of these orbitals is antibonding. One electron pair therefore finds itself in an orbital that is increasing in energy, and would, if it remained in that orbital throughout, end up in one of the o* orbitals of the product to yield the excited configuration indicated on the right in the diagram. The pattern is the one we associ- ate with a thermally forbidden process. The situation is reversed for the 72s + a4s addition, Figure 10.20 illustrates th case; now the bonding orbitals all transform directly to bonding orbitals of the prod- uct and the ymmeury imposed barrier. As with the electrocyclic processes, the correlation diagrams illustrate clearly the reason for the striking difference observed experimentally when the number of electrons is increased from four to six. The reader may verify that the 4s + 4s reaction will be forbidden. Each change of the total number of electrons by two reverses the selection rule. Figure 10.21 shows the geometry for a 72s + w2a addi gram presented in Figure 10.22 shows that the change of one component to the antarafacial mode of addition has caused a change from forbidden to allowed. When both ethylene units enter in the antarafacial manner (Figures 10.23 and 10.24), the reaction is again forbidden. It may be seen that a cycloaddition of two molecules will be thermally allowed when the total number of electrons is 4q + 2, 0, 1, 2... ce if both components are suprafacial or both antarafacial, and when the total number of electrons is 4g, q = 0, 1, 2... . , if one component is suprafacial and one antarafacial. ion. The correlation dia ‘The Generalized Woodward—Hoffmann Pericyclic Selection Rules Comparison of the Woodward—Hoffmann orbital correlation method with the Dewar—Zimmerman aromaticity method shows that the reacting molecular orbitals of a correlation diagram are constructed from exactly those basis orbitals that make up the interaction diagram of the transition state. That interaction diagram can be thought of as being built up from components, where a component is a fragment corresponding to one of the molecular orbital systems used in the Woodward— Hoffmann correlation diagram. A component, then, may be defined as « fragment corre- sponding to any subdivision of an interaction diagram. Ordinarily, components are chosen Figure 10.20 Symmetry and correlation diagram for the 725 + as cycloaddition. Because no ymmetry element transforms orbitals of one molecule into those of the other, ethylene and butadiene orbitals may be considered separately. The intended crossings of orbitals of the same symmetry (dashed lines) are avoided, so that the actual correlations are as indicated by the solid lines, The relative energies of the orbitals are not depicted quantitatively. The energy order on the left side of the diagram is that given by the Hiickel method (Chapter 1, Appendix 1) for the a orbitals of ethylene and butadiene; on the right side the order is based on the assumption that a small through-space interaction across the ring will place the & orbital symmetric combinations slightly lower than the antisymmetric combina- tions. 10.5 Correlation Diagrams 883 (x iG gas 8 ¢ CON Sec SC U1 Hef See Symmetry £ Symmetry aK vo ste Kp 884. THe Turory or Pericyetic Reactions G © Figure 10.21 ‘The geometry of approach for the 725+ m2a cycloaddition, A Cz axis of rotation is maintained throughout. to correspond to a set of orbitals of a particular bond system, but a component could be any arbitrarily defined fragment of the interaction diagram. The other require- ments we impose for a component are that the basis orbital signs be chosen in such a way that he interaction path within the component connects only lobes of like sign and that the mumber of electrons a component carries is even. (This latter restriction is possible because we are considering processes for which the total number of electrons is even.) With a component defined as described above, we can now state concisely what we mean by suprafacial and antarafacial components: We illustrate these ideas by constructing interaction diagrams for some of the reactions for which we have examined correlation diagrams. Look first at the cyclobutene—butadiene electrocyclic reaction. Considered in the ring-opening direc- tion it is a 72 + 62 process. The disrotatory opening (Equation 10.39) is 2s + a2s: 10.5 Correlation Diagrams 885 Symmetry elements y Reactants Product as aro” " A aie Molecular Symmetry Symmetry Molecular orbitals G CG orbitals Energy A cy Figure 10.22 Orbital correlation diagram for the 72s + 72a cycloaddition. Reactants are ar- ranged as in Figure 10.21(a), with p orbitals on the more remote ethylene unit seen end-on. The product corresponds to Figure 10.21(c). See the caption of Figure 10.19 for an explanation of the relative energies of orbitals on the left side of the diagram. 886 Tue THrory oF Perteycuic Reactions: @ © Figure 10.23 The geometry of approach for the 72a + 72a cycloaddition. The C2 axis is mid- way between the two molecules tee . 028) or % Components are 2 43 10.5 Correlation Diagrams 887 Symmetry elements £ a : |=. ‘ id \ ’ 3 ae Molecular Symmetry Symmetry Molecular orbitals C G orbitals Energy Figure 10.24 Orbital correlation diagram for the 72a + 2a cycloaddition. The forms of the molecular orbitals indicate a correlation according to the dashed lines. The in- tended crossings of orbitals of the same symmetry are avoided, and the actual correlations are according to the solid lines. See the caption of Figure 10.19 for an explanation of the relative energies of orbitals on the left side of the diagram, 888 Tie THEoRY oF PeRicycLiG REACTIONS Construct the interaction diagram by joining the components according to the interac tion path established by disrotatory opet a LY ) a2 44 Both components are suprafacial. ‘The interaction diagram shows a Hitckel ring of four electrons; the process is thermally forbidde ‘The conrotatory opening (Equation 10.40) has components as shown in 45; the conrotatory (10.40) reaction establishes the interactions shown in 46. ‘The process is 72s + @2a; the ring is anti-Hiickel, and the reaction is thermally allowed for four electrons. We could just as "8 g * “ Y) 45 46 well have called this reaction 72a + 2s, as shown in 47. The resulting interaction diagram still has one phase inversion and is entirely equivalent to 46. ae 47 The 72s + 74s cycloaddition (Figure 10.20) would have components as illustrated in 48, from which we derive interaction diagram 49. The interaction diagram shows no phase inversion; a Hiickel transition state of six electrons is aromatic and the 10.5 Correlation Diagrams 889 PR GY 43 RS 48 49 reaction is thermally allowed. We could have subdivided the orbital systems further by considering each of the double bonds of butadiene as a separate component (50) and joined the three components in a 72s + #2s + 2s process (51). Our conclusion that 3B be Y 50 51 the reaction is thermally allowed would have been unchanged even if we had set up the diagram as shown in 52 or 53; the two phase inversions introduced by an awkward but permissible choice of signs are equivalent to no phase inversions, and the dia- grams still show Hiickel rings. EH FY ae Se 52 53 The forbidden or allowed nature of the reaction will be determined by the suprafacial—antarafacial properties of the components and the number of electrons carried by each. We have seen that each addition of two electrons to the perieyclic ring reverses the selection rule, as does change of the faciality of a component. A general- ized rule can thus be formulated in terms of the classification of the components according to two criteria: whether they are supra- or antarafacial, and whether they 890 ‘THe THroRy oF PeRtcycLic Reactions: contain 4q + 2 or 4r electrons, q and r being any integers including zero. Woodward and Hoffmann established in this way a general pericyclic selection rule:? A simplified statement of the rule is possible if the interaction diagram is always consid- ered to be built from two-electron components, so that there are no components of the (47) type! It can be shown that the correspondence between the Woodward—Hoffmann selection rules and the Dewar-Zimmerman aromaticity rules illustrated in the few examples cited above is general. The two statements of the rules for pericyclic reac- tions are equivalent.*! Complex Systems The reactions we have considered all have interaction diagrams consisting of a single closed loop of orbitals. Such diagrams are said to be simply connected. Non-simply connected systems have more than one ring in the interaction diagram. The Wood- ward—Hoffmann and Dewar—Zimmerman rules apply only to simply connected sys- tems. Reactions that are not simply connected are sometimes difficult to analyze. Dewar’s perturbation theory, which we illustrated in Section 10.3, provides a method for assessing energy changes that occur when extra cross-ring interactions are intro- duced into a cyclic conjugated system."* Woodward and Hoffmann have shown how to find solutions in complex systems by tracing correlations of orbitals from starting materials to products by conserving the nodal structure when there are no symmetry elements bisecting the bonds being broken and formed.*’ They have analyzed the benzene-prismane example in Equations 10.41-10.42. The reaction is formally a ms + 72a + m2a “Woodward, R. B.: Hoffinann, R. “The Conse:vation of Orbital Symmetry”: Verlag Chemie: Weinheim/Bergstr,, Ger many, and Academic Press: New York, 1970, p. 169. It is understood that reactions forbidden in the ground state are allowed in the excited stare and vice versa. 3G) Day, A.C. J. Am. Chem. Soc. 1975, 97, 2431; (b) Lowry, T- HL; Richardson, K. . “Mechanism and Theory in Organic Chemistry”; Ist ed; Harper & Row: New York, 1976, p. 611. S Dewar, M. J. 8. “The Molecular Orbital Theory of Organic Chemistry"; McGraw-Hill; New York, 1969; Chap, 6. Woodward, R. B.; Hoffinann, R. “The Conservation of Orbital Symmetry”; Verlag Chemie: Weinheim/Bergstr., Ger- many, and Academic Press: New York, 1970, p. 107. 10.6 Correlation of Electronic St 891 ~ (10.41) Z I (10.42) cycloaddition (84) and should be thermally allowed. The correlation diagram, how- ever, shows that the symmetry imposes a correlation between bonding and antibond- iig levels, isd s0lthe thetmal reaction ts achially forbiddeny Experlinentally Otlsizs found that the isomerization of hexamethylprismane to hexamethylbenzene (Equa- tion 10.48) is exothermic by 91.2 kcal mol! (382 k] mol~!).*! Despite this large ot cy favorable energy change, the prismane rearranges by way of intermediate structures (Equation 10.44) with the thalpy for the direct conversion (Equation 10.43) must be greater than 33 kcal mol (138 kJ mol~!). This barrier is remarkably high for a reaction so strongly exothermi the conclusion that there is a symmetry imposed barrier scems justified. ctivation enthalpies indicated, Hence the activation en- 1 (10.44) 10.6 GORRELATION OF ELECTRONIC STATES The Woodward—Hoffmann pericyclic est in the pathways of forbidden reactions and of excited-state proce action theory has generated substantial inter- ses, beginning Oth, |. FM. Angew Chem. Int Ed, Engl, 1968, 7, 646; (b) Oth, J. BoM, Rech, Trav, Chin, 1968, 87, 1185. 892 THe THEORY oF PeRicycLic Reactions with a paper by Longuet—Higgins and Abrahamson”? that appeared simultaneously with Woodward and Hoffmann’s first use of orbital correlation diagrams.*® We have noted in Section 10.5 that the orbital correlation diagram predicts that if a forbidden process does take place by a concerted pericyclic mechanism,”’ and if electrons were to remain in their original orbitals, an excited configuration should be produced. Similarly, if reactants start out in a singly excited configuration (Figure 10.17), the diagrams imply that one elecrtron will decrease in energy and the other increase, the result being a singly excited product configuration. A number of investigators have pointed out that under these circumstances the electron pairs are ordinarily not ex- pected to remain in their original orbitals.” In Section 1.2 we discussed the electronic configuration and the electronic state. Recall from that discussion that a configuration assigns the electrons to the orbitals of a molecule to give a first approximation to an electronic state; configuration intera tion then mixes configurations and leads to better descriptions of states. We must now add to the earlier treatment a description of how symmetry affects the interaction of configurations. Figure 10.25 shows the same orbitals and configurations as Figure 1.10, but with symmetry designations added.*" The symmetry behavior of each elec- tron is determined by the symmetry of its molecular orbital; the symmetry of the configuration as a whole is found by taking the product of symmetries of all the electrons." Thus to find the symmetry of a configuration, each electron in a symmet- ric orbital is given a symmetry value of +1 and each electron in an antisymmetric orbital a symmetry value of —1; the symmetry of the configuration is the product of all the electron symmetries. For example, a configuration with only two electrons, one of which is in a symmetric orbital and the other in an antisymmetric orbital, will be antisymmetric [(+1) x (—1) = ~ 1], while a pwo-electron configuration with both elec- trons either in symmetric orbitals or in antisymmetric orbitals will be symmetric [(F1) x (41) = +1 or (1) x (1) = 41, respectively]. When configurations interact, symmetry properties govern the interaction in the same way as they govern the inter- actions of orbitals: only configurations of the same symmetry interact. Configuration interaction plays a relatively minor role in the description of ground-state molecules, except in ab initio calculations where high accuracy is desired, because the ground configuration is normally well separated from other configura- *Longuet-Higgins, H. Cz Abrahamson, E, W. J. Am. Chen. Soe. 1965, 87, 2045, *Hoflmann, Ri Woodward, R. B,J. Am. Clem. Soe. 1965, 87, 2046, 7 stepwise pathway, not subject the same symmetry restrictions, is always possible. 3) See note 35. See also, for example, (b) Dewar, M. J. S. Angew Chem. fut. Fd. Engl. 1971, 10, 761; (c) van der Lug W.Th. A.M; Oosterhoff, L. J.J. Am. Chem, Soc, 1969, 97, 6042; (a) Dougherty, R. C.J- Am. Chem, Soc. 1971, 92, 7187; (@) Baldwin, |. Ez Andrist, A. Hy; Pinschmidt, Js. R. K. Ace. Chem. Res. 1972, 5, 402; f) Hoffmann, R.; Swaminathan, S ‘Odell, B. Ge; Gleiter, RJ. tn. Chem. Soc. 1976, 92, 7091; (g) Epiotis, N. D.J. Am. Chew. Soc. 1978, 95, 1191, 1200, 1214; (h) Fpiotis, N.D.! Shaik, . J. Am. Chem. Soe. 1978, 100, 1, 9, 18, 29; () Gerhart, W.; Poshusta, R. Ds Mich, J Am. Chem. Soc, 1976, 98, 6427; () Williams, D. HL Acc. Chem. Is. 1977, 10, 280, Our configuration specification in Figure 10.25 is still oversimplified. We have omitted specification of electron spin and have also shown the orbital energies as though they were independent of occupancy, an approximation that does not hold in the self-consistent-ield (SCF) molecular orbital theories that are essential for adequate treatment of excived states Furthermore, the configurations should be given in the form of Slater determinants, as described in the Appendix to Chapter 1 “The discussion is restricted 10 molecules complex symmetry designations arise wh th axes of rotational symmetry of order no higher than wofold. More » a rotation axis of order th 1e oF greater is present suonperann woneInByt0> fq pauresqo suonemSryM09 Jo suoneuiquio> snowea at) ade sans otWONDA[a ALLO ay], “S8e] DEIAIUE OS puR KayatUIAs ay] Jo S1OIPO WoL, ADIpAEy Ie Up pur Of suonemnSijuo; A pue %A suoneANS1ju09 arp 10J parexsnyp st uoNDeIa WUT UONANBIUOD & ons ‘1ySL1 wonog ayp ry “XIU puk (Op sjequo se isnf) 1>eLaiMT yas Aayp “ABx9u9 m1 asp axe ArnauTUAS auIes JY JO suONENSyUOI OM FAAdUDYM “SATIS OUONDaLa ay}. oO} SUORUNXoAdde js11} P219pIsuOd aq deur suonesn3yuo> ay, “(Y 40} [— ‘§ 10} 1+) suOIDa[> amp Jo sainouUds app SuKidnjmLE YpiM ‘suonNDa|a ayy 10J SHONdUNy [eGo Jo ImNpord & s1 VoNdUNy uonemijuoD Yprg “a[es ABU uP uO UMOYs D1 suONeINSyUOD dP “Yo! ‘Mojog “uoneuxosdde siup ur uonemSiyuoo y>ea 10} oures aup s119s [eIIgI0 se[Nd9JoW dup ‘suONeINByuo> snoLtwA 10) sa;uEd -n290 [eGo areNsNy! pur soniodosd AnouMUAs Ayoods susesBeIp foaof-AB10U9 penquo ‘2aoqy WaUI9p Anawtuxs suo yt ano9[our eonotpodAy ¥ Jo soveis paHxd soy ¥ pu oyeIs punosB ouoIDa|a 3yT ezoT aun 893 10.6 Correlation of Electronic States a ee i $s ¢0,0n,09- 0% ipaitiy fo SL, fe faye tas A ‘sas0ug PEMD = Es = S§ $n cent) (10) = tae Sa 2 ee suopemsyuoy oe th sts i 4 rt y—— 4 spuog sa sy a eK . ‘6 vee rt — v—— ‘ af s “0 s——* s Sq Suipuognuy v—"* ym ym stenaio seino2t0n8 —_ $$ 894 The THeory oF Pericyciic Reactions tions in energy and so does not interact strongly with them. ‘Therefore the ground configuration is usually a good approximation for the ground state. But, as we have already seen in earlier descriptions of reactions according to the valence-bond model of Pross, configuration interaction can become very important in determining reac~ tion coordinate profiles, because as the molecules distort and interact the configura- tion energies change and configurations originally far apart may come close ergy. These ideas also provide a framework for looking at pericyclic reactions. n en- Ground-State Allowed, Excited-State Forbidden Reactions Figure 10.26 illustrates the #2s + 4s cycloaddition from the point of view of clec- tronic states. We are making the approximation here that the three states we are interested in can, in reactant and in product, be represented by single configurations. Note that the ground state, represented by configuration W,, goes over to product ground state (E,) with no symmetry imposed barrier; hence the reaction is thermally allowed. Reference to the energy behavior of the individual orbitals for this reaction (Figure 10.20) shows, on the other hand, that if one were to start from the excited state represented by configuration W,, keeping all electrons in their original orbitals, ‘one would end up in the higher energy state represented by configuration &. Thus there is an intended correlation bewween ¥, and &.. Part way along the path another state of the same symmetry, represented by configuration V,-, approaches , in energy because it intends to correlate to the lower configuration &. The resulting intended crossing is shown by the dotted lines in Figure 10.26. ‘This intended crossing never actually occurs because when W, and W,- come close in energy configuration interac- tion takes over. The two configurations lose their separate identities as the lower energy configuration , mixes into itself more and more of VW, and vice versa. The result is that if the molecule starts out in Y,, it follows the solid line and emerges in &, having climbed the energy hill to a height close to the intended crossing point. Be- cause the orbital symmetry correlations have imposed this energy barrier, the reaction is said to be symmetry forbidden in the first excited state. Excited-State Allowed Reactions In Figure 10.19, which shows orbital correlations for the thermally forbidden 72s + 725 cycloaddition, we saw that by raising one electron from orbital 7» to orbital 7% to form a configuration approximating the first excited state, the energy barrier to the ground-state cycloaddition could be overcome. The penalty paid by one electron dimbing to higher energy is compensated by the other electron falling to lower en- ergy. Figure 10.27 shows an approximate state correlation diagram for this process. The first excited state, represented by configuration V,, correlates to the first excited state of the product, represented by &. There is no symmetry imposed barrier along the path, so the reaction is allowed in the first excited state in agreement with experi- ment. Also shown in Figure 10.27 are the correlations of the ground configurations and the doubly excited configurations, V,, é,, W,, and &. The latter two configurations represent approximately an excited state obtained by promoting two electrons from the highest occupied orbital of reactant or product respectively to the lowest unoccu- pied orbital. (Refer to Figure 10.19 for the ¢ correlations.) Because configura-

You might also like