The principle of nonperfect synchronization: recent developments

The principle of nonperfect synchronization: recent developments CLAUDE F. BERNASCONI Department of Chemistry and Biochemistry, University of California, Santa Cruz, CA 95064, USA 1 Introduction 223 Intrinsic barriers 224 Transition state imbalance and the PNS 224 Scope of this chapter 225 2 Proton transfers in solution 226 The effect of resonance on intrinsic barriers and transition state imbalances 226 Why does delocalization lag behind proton transfer? 237 Other factors that affect intrinsic barriers and transition state imbalances 238 3 Proton transfers in the gas phase: ab initio calculations 261 The CH3Y/CH2=Y Systems 261 The NCCH2Y/NCCH=Y Systems 280 Aromatic and anti-aromatic systems 282 4 Other reactions 293 Nucleophilic additions to alkenes 293 Nucleophilic vinylic substitution (SNV) Reactions 298 Nucleophilic substitution of Fischer carbene complexes 303 Reactions involving carbocations 309 Miscellaneous reactions 312 5 Summary and concluding remarks 316 Acknowledgments 319 References 319

1

Introduction

This year marks the 25th anniversary of the principle of nonperfect synchronization (PNS); it was introduced in 19851 as the principle of imperfect synchronization (PIS) but in later papers and reviews2–4 the name was changed due to the awkwardness of the acronym PIS. The foundations of the PNS rest mainly on a marriage between two fundamental concepts of physical organic chemistry, i.e., the concept of intrinsic barriers and that of transition state imbalances.

223 ADVANCES IN PHYSICAL ORGANIC CHEMISTRY VOLUME 44 ISSN: 0065-3160 DOI: 10.1016/S0065-3160(08)44005-4

 2010 Elsevier Ltd. All rights reserved

224

C.F. BERNASCONI

INTRINSIC BARRIERS

Marcus5–8 taught us that the most appropriate
and useful kinetic measure of chemical reactivity is the intrinsic barrier DG‡o rather than the actual barrier (DG‡), or the intrinsic rate constant (ko) rather than the actual rate constant (k) of a reaction. These terms refer to the barrier (rate constant) in the absence of a thermodynamic driving force (DGo = 0) and can either be determined by interpolation or extrapolation of kinetic data or by applying the Marcus equation.5–8 For example, for solution phase proton transfers from a carbon acid activated by a p-acceptor (Y) to a buffer base, Equation (1), ko may be determined from Br½nsted-type plots of log k1 or Bν + H

C

Y

k1

C

Y– + Bν+1

k–1

(1)

log k–1 versus log K1 (K1 = k1/k–1) by interpolation or extrapolation to K1 = 1,9 while DG‡o can be calculated from ko via the Eyring equation. Or, using the Marcus equation which, in its abbreviated form, is given by Equation (2), allows one to solve for DG‡o for a given set of DG‡ and DGo values.

‡

DG ¼

DG‡o

  DGo 2 1þ 4DG‡o

ð2Þ

The benefit of determining intrinsic barriers or intrinsic rate constants as measures of chemical reactivity is that they can be used to describe the reactivity of an entire reaction family, irrespective of the thermodynamic driving force of a particular member of that family and to make comparisons between different families. For example, DG‡o or ko determined for the deprotonation of acetylacetone by a series of secondary alicylic amines may be compared with DG‡o or ko for the deprotonation of nitroacetone by the same series of secondary amines. This comparison would provide insights into how the change of one of the pacceptor groups from acetyl to nitro may affect the intrinsic proton-transfer reactivity without regard to how this change may affect the pKa value of the carbon acid. Furthermore, DG‡o or ko for the reaction of acetylacetone with secondary alicyclic amines may be compared to DG‡o or ko for the deprotonation of the same carbon acid by a series of primary amines, leading to insights as to how differences in the solvation characteristics between primary and secondary amines may affect their intrinsic kinetic reactivity. TRANSITION STATE IMBALANCE AND THE PNS

The PNS derives from the realization that the majority of elementary reactions involve more than one concurrent event such as bond formation, bond

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION Product stabilizing factor

Reactant stabilizing factor

Develops late: ko ↓; ΔGo‡ ↑

Lost early: ko ↓; ΔGo‡ ↑

Develops early: ko ↑; ΔGo‡ ↓

Lost late: ko ↑; ΔGo‡ ↓

Product destabilizing factor

Reactant destabilizing factor

Develops late: ko ↑; ΔGo‡ ↓

Lost early: ko ↑; ΔGo‡ ↓

Develops early: ko ↓; ΔGo‡ ↑

Lost late: ko ↓; ΔGo‡ ↑

225

Chart 1

cleavage, solvation/desolvation, transfer and delocalization/localization of charge, etc., and often these processes have made unequal progress at the transition state. When this is the case, the reaction is said to have an imbalanced transition state, a term introduced and popularized by Jencks,10,11 although others before him had recognized this phenomenon in various reactions, especially in E2-eliminations.12–14 The virtue of the PNS is that it establishes a connection between transition state imbalances and intrinsic barriers of reactions. Its original formulation is still valid today; it states that any product stabilizing factor whose development lags behind the main bond changes at the transition state, or any reactant stabilizing factor whose loss at the transition state is ahead of these bond changes, increases the intrinsic barrier or decreases the intrinsic rate constant. For product stabilizing factors that develop ahead of the main bond changes, or reactant stabilizing factors whose loss lags behind the bond changes, the effects are reversed, i.e., there is a decrease in DG‡o or an increase in ko. For product or reactant destabilizing factors the opposite relationships hold. Chart 1 provides a summary of these various manifestations of the PNS. Product or reactant stabilizing factors that have been studied thus far include resonance/charge delocalization, solvation, hyperconjugation, intramolecular hydrogen bonding, aromaticity, inductive, p-donor, polarizability, steric, anomeric, and electrostatic effects, as well as ring strain and soft–soft interactions. Product or reactant destabilization factors are mainly represented by anti-aromaticity, steric effects in some types of reactions, and, occasionally, electrostatic effects. What makes the PNS particularly useful is that it is completely general, mathematically provable,4 and knows no exception.

SCOPE OF THIS CHAPTER

Regarding the scope of this chapter, the main focus is on work published after my detailed 1992 review.4 Older material will only be presented when necessary

226

C.F. BERNASCONI

to put new results into perspective to emphasize important points neglected in earlier reviews or to correlate new data with old results in the form of summary tables or graphs. Proton transfers from carbon acids have continued to play a particularly prominent role in illustrating the multiple manifestations of the PNS and hence their studies constitute a major part of this chapter. This is especially true for ab initio calculations of proton transfers in the gas phase that have been performed after 1992 and have added novel insights into the workings of the PNS. Also new is an important expansion of the list of product stabilizing factors to include aromaticity and, for product destabilizing factors, anti-aromaticity. Other reactions for which a discussion of their structure-reactivity behavior in terms of the PNS has provided valuable insights include nucleophilic addition and substitution reactions on electrophilic alkenes, vinylic compounds, and Fischer carbene complexes; reactions involving carbocations; and some radical reactions. The number of reports on reactions that have been discussed in the context of the PNS or that would benefit from being treated within this framework far exceeds the space available in this chapter. Hence the purpose of this review is not to give a comprehensive account of all such reactions but rather to be selective and focus on those cases that provide genuine insights into the workings of the PNS.

2

Proton transfers in solution

THE EFFECT OF RESONANCE ON INTRINSIC BARRIERS AND TRANSITION STATE IMBALANCES

In contrast to proton transfers between normal acids and bases, which typically have very high intrinsic rate constants that are close to the diffusion-controlled limit15,16 and depend little on the nature of the acid, proton transfers from carbon acids have intrinsic rate constants that vary strongly with structure and are mostly much lower than the diffusion-controlled limit.15,17,18 Table 1 summarizes DG‡o and log ko values for a number of representative examples.4,18–28 The data show a dramatic decrease in ko (increase in DG‡o ) as the p-acceptor strength of the activating groups increases from the top to the bottom of the table. The main reason for the observed trend is that the transition state is imbalanced in the sense that the degree of charge delocalization into the p-acceptor lags behind the proton transfer. This is shown, in exaggerated form, in Equation (3) (for a more nuanced representation of the transition state see below) for a ‡

Bν + H

C

Y

k1

ν+δ B H

δ– C Y

k–1

C

Y– + BHν + 1

(3)

Entry 1

Solvent, Ta

C–H acid CH2(CN)2

H2O, 25C

50% DMSO, 20C

2 H

3

pKCH a

log kob

11.2

7.5

DG‡o kcal mol1

References

7.2

18

9.53

4.58

10.9

19

11.54

4.03

11.7

20

50% DMSO, 20C

4.70

3.90

11.9

4

50% DMSO, 20C

12.62

3.70

12.1

21

50% MeCN, 25C

12.50

3.70

12.1

22

50% DMSO, 20C

6.35

3.13

12.9

22

50% DMSO, 20C

9.12

2.75

13.4

24

H2O, 20C

7.72

2.29

14.0

20

CN

CH3

PhSO2CH2 CH3

OOC

CH3

OOC

CH2

4

5

O2N

6

(CO)5Cr

CH2CN

H2O, 20C

OMe CH3 CO CH2 CO

7 O

8

C

O

CH3CCH2CCH3

PhCCH2

+ N CH3

227

O

9

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

Table 1 Representative intrinsic rate constants (ko) and intrinsic barriers ( DG‡o ) for the deprotonation of carbon acids by secondary alicyclic amines

228

Table 1 (continued ) Entry

C–H acid

10

(CO)5Cr

11

CH3NO2

12

+ Cp*Ru

13

PhCH2NO2

Solvent, Ta

pKCH a

log kob

DG‡o kcal mol1

50% MeCN, 25C

10.40

1.86

14.6

50% DMSO, 20C

11.32

0.73

16.1

25

50% DMSO, 25C

5.90

0.10c

17.2

26

50% DMSO, 25C

7.93

–0.25

17.7

25

22

OMe C CH2Ph

CH2NO2

References

2,4-(NO)2C6H4 CH2

14

50% DMSO, 25C

10.9

–0.55

18.1

27

2,4-(NO)2C6H4

15

CH3NO2

H2O, 20C

10.28

–0.59

17.9

25

16

PhCH2NO2

H2O, 20C

6.77

–1.22

18.7

25

H2O, 25C

2.50

–2.15

20.3

28

CH3

N+ O N

17 NO2 a

50% DMSO = 50% DMSO–50% water (v/v); 50% MeCN = 50% MeCN–50% water (v/v). In units of M1 s1. c Reaction with primary aliphatic amines. b

C.F. BERNASCONI

O 2N

O–

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

229

generalized carbon acid with the p-acceptor Y. Because of the imbalance, the transition state derives only a minimal benefit from the stabilizing effect of charge delocalization into the p-acceptor, and this is the reason why the intrinsic barrier is high. It is as if at the transition state the carbon acid were less acidic than indicated by its pKa. The increase in DG‡o is perhaps even more easily understood when the reaction in the reverse direction (k–1) is considered. Here the reason for the high intrinsic barrier is that most of the resonance stabilization of the carbanion is lost at the transition state, i.e., it costs extra energy to localize the charge on the carbon in reaching the transition state. The increase in DG‡o with increasing acceptor strength thus reflects the increasing cost of localizing the charge and goes hand in hand with an increase in the imbalance. In cases where there is strong solvation of the carbanion, as for example hydrogen bonding solvation of enolate or nitronate ions in hydroxylic solvents, the intrinsic barrier is increased further because the transition state cannot benefit significantly from this solvation. This is the reason why DG‡o for the deprotonation of nitroalkanes in water is particularly high, i.e., much higher than in dipolar aprotic solvents, see, e.g., entry 11 versus 15 and entry 13 versus 16 in Table 1. These solvation effects will be discussed in more detail below. Evidence of imbalance based on Br½nsted coefficients. A. aCH > bB Since the observed trends in the intrinsic barriers can plausibly be explained by assuming that charge delocalization lags behind proton transfer (or charge localization is ahead of proton transfer in the reverse direction), this may be taken as evidence for the existence of imbalance. Nevertheless, independent evidence for the presence of transition state imbalances would be desirable. Such evidence exists in the form of structure-reactivity coefficients such as Br½nsted a and b values and has in fact been known before the connection between imbalance and intrinsic barriers was recognized. For example, assume that one of the groups attached to the carbon is an aryl group with various substituents Z [Equation (4)] and that the transition state is imbalanced in the same way as shown in Equation (3). One may, for a given B , determine a Br½nsted a value designated

‡

Bν

k1 + H

C

Z

Y

ν+δ B H

δ– Y C

C

k–1

(4) Z

Z

Y– + BHν + 1

230

C.F. BERNASCONI

as aCH (CH for carbon acid) by measuring k1 as a function of the acidity constant of the carbon acid as varied by changing Z, and, for a given Z, determine a Br½nsted b value designated as bB by measuring k1 as a function of the basicity of B . As long as B is not a carbanion with its own resonance stabilization, it has generally been assumed that bB is an approximate measure of the degree of proton transfer at the transition state.11,29,30 However, with aCH the situation is different when there is an imbalance, i.e., aCH is not a good measure of proton transfer. Rather, aCH is typically larger than bB. This is because, due to the closer proximity of the charge to the Z-substituent at the transition state compared to that in the product ion, the substituent effect on the transition state is disproportionately large relative to that on the carbanion. This means that the sensitivity of k1 to the substituent Z is disproportionately strong compared to that of the acidity of the carbon acid and hence aCH ¼ d log k1 =d log kCH is exalted, a i.e., aCH > bB. Note that for the Br½nsted coefficients determined in the reverse direction, bC ¼ d log k1 =dpKCH and aBH ¼ d log k  1 =d log KBH a a , the relationship bC < aBH holds. This is a consequence of the equalities aCH þ bC ¼ 1 and bB þ aBH = 1. Table 2 summarizes aCH and bB values for some representative proton transfers where the imbalance leads to aCH > bB;31–39 below we will discuss cases where the imbalance leads to aCH < bB. The best-known and one of the most dramatic examples of an imbalance is provided by the deprotonation of arylnitromethanes by secondary alicyclic amines in aqueous solution (entry 9 in Table 2) where aCH = 1.29 and bB = 0.56.31 In this case the imbalance is so large that aCH is greater than the boundary value of 1.0. This implies that in the reverse direction bC is negative (–0.29) which means that electronwithdrawing substituents enhance not only the rate of deprotonation (k1) but also the rate (k–1) of the protonation of the carbanion. The large magnitude of the imbalance as reflected in the large difference between aCH and bB (aCH – bB = 0.73) is the result of the exceptionally strong p-acceptor strength of the nitro group, coupled with the strong hydrogen bonding solvation of the nitronate ion in aqueous solution. This solvation reduces the need for stabilization of the nitronate ion by the Z-substituent and hence decreases the dependence of the acidity constant on Z. But since the transition state does not significantly benefit from the solvation, the dependence of the rate constant on Z changes little with the solvent and hence aCH becomes larger. We see again the direct connection between imbalance and DG‡o at work, i.e., the exceptionally large imbalance for the nitroalkanes goes hand in hand with the exceptionally high intrinsic barrier (see Table 1). A very large imbalance is also seen for the reaction of ArCH2NO2 with HO (entry 11); even though no bB value could be determined for this reaction to provide an approximate measure of proton transfer at the transition state, the mere fact that aCH > 1 demonstrates the presence of a strong imbalance. The same is true in even more dramatic fashion for the HO-promoted

Entry 1 2

C–H acid

Base

ArCH2CN ArCH2CH(CN)2

Solvent, T

aCH

bB

aCH – bB

References

Ar0 CH2NH2 RCOO

DMSO, 25C H2O, 25C

0.74 0.98

0.61 0.83

0.13 0.15

32 33

O

3

Z

RCOO

H2O, 25C

0.78

0.54

0.24

34

4 5 6 7

ArCH2NO2 ArCH2CH(COMe)COOEt ArCH2NO2 ArCH(CH3)NO2

Ar0 COO RCOO Ar0 COO R2NH

MeCN, 25C H2O, 25C DMSO, 25C H2O, 25C

0.82 0.76 0.92 0.94

0.56 0.44 0.55 0.55

0.26 0.32 0.37 0.39

35 33 36 31

R2NH

50% DMSOa, 25C

0.87

0.45

0.42

37

R2NH Ar0 COO HO

H2O, 25C MeOH, 25C H2O, 25C

1.29 1.31 1.54

0.56 0.50 –

0.73 0.81 large

31 38 31

HO

H2O, 25C

>>1

–

v. large

39

Z

8

CH2

O2N

NO2

NO2

9 10 11

ArCH2NO2 ArCH2NO2 ArCH2NO2 MeO

12

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

Table 2 Br½nsted coefficients and imbalances aCH – bB for the deprotonation of C–H acids by amines and carboxylate ions. Cases where aCH > bB

CH3

H

N+

C

CH3 S

50% DMSO–50% water (v/v).

231

a

232

C.F. BERNASCONI

deprotonation of the 2-benzylthiazolium ion shown as the last entry in Table 2. The authors39 report that with Z = (CH3)3Nþ the rate constant is about 1100fold higher than with Z = H while the pKCH values of the two compounds a differ by less than one half of a log unit, implying an aCH value  4. The reason why aCH is so large in this case must be related to the fact that the negative charge that builds up at the transition state [Equation (5)] is not only delocalized away from the substituent Z but completely disappears in the product, rendering the pKCH very insensitive to Z. a ‡ δ– HO H

OCH3 C

H

δ– OCH3 CH3 C N+

CH3

N+

HO– + Z

S

CH3

CH3

S

Z

(5)

OCH3 CH3

C N Z

S

+ H2O CH3

Coming back to the nitro compounds we note that in several cases the imbalances are not nearly as large as for the reactions of ArCH2NO2 with amines or HO in water. For example, for the reactions of ArCH2NO2 with benzoate ions in DMSO (entry 6) and MeCN (entry 4), the imbalances are much less dramatic than in water and are the result of smaller aCH values. This reduction is due to the strongly reduced solvation of the nitronate ion in dipolar aprotic solvents. There is also a reduction in the imbalance of the reaction of ArCH(CH3)NO2 (entry 7) with amines in water relative to the corresponding reaction of ArCH2NO2 (entry 9). This is again the result of a decrease in aCH and can be explained in terms of reduced charge delocalization into the nitro group of the nitronate ion due to steric hindrance of coplanarity by the methyl group. Evidence of imbalance based on Br½nsted coefficients. B. aCH < bB In all examples listed in Table 2 the lag in the charge delocalization behind proton transfer leads to aCH being greater than bB. There are, however, cases where the opposite is true, i.e., aCH < bB. Representative examples40–43 are reported in Table 3. Does this mean that the imbalance is reversed in these

Entry

C–H acid

B

Solvent, Ta

aCH

bB

aCH – bB

References

R2NH

90% DMSO, 20C

0.46

0.64

–0.18

40

R2NH

50% DMSO, 20C

0.29

0.49

–0.20

41

RND2

5% DMSO, 25C

0.46

0.70

–0.24

42

RNH2

H2O, 25C

0.27

0.55

–0.28

43

NO2

1

Z

CH2CN

2

+ PPh3

H

+ Z S(CH3)2

3

O

4 a

Z

H

PhCCH2

+ N CH2Ar

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

Table 3 Br½nsted coefficients and imbalances aCH – bB for the deprotonation of C–H acids by amines. Cases where aCH < bB

90% DMSO = 90% DMSO–10% water (v/v); 50% DMSO = 50% DMSO–50% water (v/v); 5% DMSO = 5% DMSO–95% D2O (v/v).

233

234

C.F. BERNASCONI

reactions; in other words, is charge delocalization ahead of proton transfer? Closer inspection of the situation illustrated with the example of Equation (6) demonstrates that this is ν +δ B H

CH2CN

δ– CH

NO2

‡ CHCN

CN

NO2

NO2

Bν +

Z

+ BHν + 1

(6)

Z

Z

not the case. The reason why aCH < bB is that here the lag in the charge delocalization creates a situation where it is the charge in the product ion that is closer to the substituent than the developing charge at the transition state; this is opposite to the situation in Equation (4) or all examples of Table 2. This makes aCH disproportionately small and leads to aCH < bB. The last entry in Table 3 is of particular interest because there is potential competition between two p-acceptors stabilizing the product. There is evidence indicating that resonance

O– Ph

C

O + N CH2Ar

CH

Ph

C CH

a

N

CH2Ar

b

structure b is dominant. Hence the reaction can be represented by Equation (7) which shows that at the ‡ O

Ph C

CH2

O

Ph ν +δ B

C H

δ– CH

Ph

O C CH

(7) Bν +

+ BHν + 1 N+

N+

N

CH2Ar

CH2Ar

CH2Ar

transition state the negative charge is far away from the aryl group but moves closer to it in the product, thereby neutralizing the positive charge.

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

235

The above interpretation of the imbalance is supported by a comparison of 44 aCH values for the deprotonation of 1 and 2 by HO and CO2 For the 3 .  reactions with HO , aCH(1) = 0.59

+ N

ArSO2CH2

+ N

PhSO2CH2

CH3

CH2Ar

2

1

and aCH(2) = 0.33, while for the reactions with CO2 3 , aCH(1) = 0.45 and aCH(2) = 0.29. The aCH(2) values are seen to be quite small for the same reason as in the reaction of Equation (7), i.e., because the product is dominated by a resonance structure that is analogous to that in Equation (7). On the other hand, aCH(1) is larger than aCH(2) because in this case the aryl group is on the other side of the molecule and hence closer to the developing negative charge at the transition state. Other examples of imbalances The determination of Br½nsted coefficients provides the most transparent tool for the evaluation of imbalances, but there are other ways to probe transition structure in search for evidence of transition state imbalances. Terrier et al.45 reported that the change to a more electron-withdrawing Zsubstituent in 3

δ– OMe X

X H

H O2N

C

NO2

Z

O2 N

C

NO2

Z Y

3

δ–

Y 4

increases the rate of deprotonation by MeO more than making X or Y more electron withdrawing, but the change in X or Y enhances the thermodynamic acidity of 3 more than the change in Z. This finding is consistent with the imbalanced transition state 4. Pollack’s group46 has studied the deprotonation of substituent 2-tetralone by HO. Based on a combination of kinetic data and 13C NMR spectra they

236

C.F. BERNASCONI

estimated the charge distributions in the transition state and anion as shown in 5 and 6, respectively. These charges imply a highly imbalanced transition state with a strong lag in the delocalization of the charge into the carbonyl group. –0.45 OH –0.33

H –0.26 –0.29 O –0.03

O

–0.68

–0.08

–0.03

0.09

5

6

Based on secondary kinetic deuterium isotope effects and some ab initio calculations, Alston et al.47 calculated that the transition state in the deprotonation of acetaldehyde by HO is imbalanced in the sense shown in Equation (3). Anslyn’s48 group examined the effect of the phenolic OD group on the deprotonation of 7 by imidazole to form the enolate ion 9. This OD group leads to a significant stabilization of the ‡

D OD

O C

B +

O

O C

CH3

D O

δ– CH2

O– C

CH2

+ BH+

(8)

H Z 7

Z

B 8

δ+

Z 9

enolate ion 9 due to intramolecular hydrogen bonding, but its effect on the rate of deprotonation is quite small. The authors concluded that intramolecular hydrogen bonding at the transition state (8) is only minimally developed because there is only a small amount of charge on the carbonyl oxygen while most of the charge resides on the carbon. Amyes and Richard49 deduced the presence of a transition state imbalance in the deprotonation of methyl and benzylic mono carbonyl compounds by HO from the linearity of the Br½nsted plot of the rate constants versus the pKa of these carbonyl compounds. They argued that because of the large reactivity range the Br½nsted plot should have shown ‘‘Marcus curvature’’5– 8 if the intrinsic barriers for these reactions were all the same and hence the absence of such curvature indicates changes in the intrinsic barriers. They

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

237

attributed the variations in the intrinsic barriers mainly to the lag in enolate ion resonance development. However, Angelini et al.50 reached somewhat different conclusions regarding the linearity of a similar linear Br½nsted plot which were based on an examination of inductive and steric effects competing with resonance effects. According to their analysis, resonance effects play a minor role while the inductive and steric effects are dominant, a conclusion supported by additional analysis based on kinetic data for the deprotonation of 2-nitrocyclohexanone.

WHY DOES DELOCALIZATION LAG BEHIND PROTON TRANSFER?

In view of the fact that nature always chooses the lowest energy pathway, one may wonder why these reactions do not proceed via a more balanced transition state with more advanced delocalization, which would presumably lower its energy. This question has been discussed at considerable length in our 1992 review4 and hence only an abbreviated version is presented here. The reason why delocalization is not more advanced is that there are constraints imposed on the transition state that prevent extensive delocalization. This was first pointed out by Kresge51 in the context of the deprotonation of nitroalkanes, but it applies to any proton transfer from carbon. The situation is represented in Equation (9) which is a more nuanced version of Equation (3) and allows for a certain degree of charge delocalization into the p-acceptor (d Y) at the transition ‡ ν + δB

Bν + H

C

Y

B

H

–δC

–δ Y

C

Y

C

Y– + BHν + 1

(9)

state. Kresge’s argument is that d Y depends on the C–Y p-bond order as well as on the charge that has been transferred from B (d B) [Equation (10)], while the p-bond order is related to d B [Equation (11)] d Y  pb:o:  d B

ð10Þ

pb:o:  d B

ð11Þ

d Y  ðd B Þ2

ð12Þ

238

C.F. BERNASCONI

since the p-bond is created from the electron pair transferred from the base. Hence d Y is given by Equation (12); it is a small number because it represents only a fraction of a fraction. A more refined picture which takes the charge on the proton-in-flight into consideration will be presented later. However, the basic features and conclusions will remain the same. In other words, because of the constraints embodied by Equations (10–12), delocalization must always lag behind proton transfer or other bond changes in other types of reactions, i.e., there cannot be exceptions. It should be noted that the origins of the imbalance have also been discussed in the context of the valance bond configuration-mixing model proposed by Shaik and Pross.52,53 This model describes the reaction energy profile in terms of the conversion of a reactant configuration (c) into a product configuration (d) and the mixing of a third configuration (e); this latter plays a

Bν: H

C c

Y

Bν + 1 H

C d

Y–

Bν: H+ – :C

Y

e

dominant role in the transition state region. The mixing in of e confers to the transition state its carbanionic character. We prefer the Kresge model because it shows that the imbalanced character of the transition state is enforced by the constraints described by Equations (10–12) whereas the Shaik–Pross model is more a post facto explanation of the imbalance. We shall return to the question of the origin of imbalances in the section on ab initio calculations.

OTHER FACTORS THAT AFFECT INTRINSIC BARRIERS AND TRANSITION STATE IMBALANCES

There are a number of factors that affect intrinsic barriers and/or transition state imbalances. Many of these may be viewed as ‘‘derived’’ effects because they are a consequence of the imbalance caused by the presence of p-receptors, i.e., in the absence of this imbalance they would not affect the intrinsic barriers even if they affect actual barriers and equilibria. Solvation Solvation can have a large effect on intrinsic barriers or intrinsic rate constants, especially hydrogen bonding solvation of nitronate or enolate ions in hydroxylic solvents. Table 4 reports intrinsic rate constants in water and aqueous DMSO for a number of representative examples.19,20,23–25,40,54–56 Entries 1–4 which refer to nitroalkanes show large increases in log ko when

Entry

1 2 3 4 5

CH3NO2 PhCH2CH2NO2 PhSCH2NO2 PhCH2NO2 CH2(COCH3)2

log ko(RCOO)a

log ko(R2NH)a

C–H Acid H2O

50% DMSO– 50% H2O

90% DMSO– 10% H2O

H2O

–0.59 –1.16 1.02 –0.86 2.60

0.73

–0.25 2.75

3.06 2.51 4.08 1.75 3.64

2.97b

3.13

3.85

1.57

1.87

2.75

55

2.51

20

2.29

2.50

4.03

3.95

3.85

20

2.85

2.76

2.54

40

–2.10 2.89

50% DMSO– 50% H2O

90% DMSO– 10% H2O

References

–0.59 3.80

1.88 5.3

25 54 54 25 24

3.18

4.53

23

CO

6

CH2

2.64b

CO

7

PhCOCH2NO2

8

PhCOCH2

9

PhSO2CH2

+ N CH3 + N CH3

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

Table 4 Solvent effects on intrinsic rate constants for the deprotonation of C–H acids by secondary alicyclic amines and carboxylate ions at 20C

NO2 O2N

CH2CN

239

10

240

Table 4 (continued ) Entry

H2O

4.44b

11 H

log ko(RCOO)a

log ko(R2NH)a

C–H Acid

50% DMSO– 50% H2O

90% DMSO– 10% H2O

H2O

50% DMSO– 50% H2O

4.58

4.39

19

3.30

2.98

56

90% DMSO– 10% H2O

References

CN

Cr(CO)3

12 (CO)3Cr a

In units of M1 s1. 10% DMSO–90% H2O (v/v).

C.F. BERNASCONI

b

CH2

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

241

the water content of the solvent is reduced. This implies that hydrogen bonding solvation of the anion leads to particularly large increases in the intrinsic barrier. The reason for the large intrinsic barrier is that solvation of the incipient charge lags behind proton transfer, just as delocalization of the charge into the p-acceptor group lags behind proton transfer. Hence the scheme of Equation (3) may be extended to include solvation as shown in Equation (13). Again, the easiest way ‡ Bν +

H

C

Y

ν+δ B H

δ– C Y

C

Y–

H2O + BHν + 1

(13)

to understand the barrier-enhancing effect is to consider the reaction in the reverse direction. To reach the transition state, energy is required not only to localize the charge on the carbon but also to desolvate the carbanion. In other words, there is a solvational PNS effect that is superimposed on the resonance/ delocalization PNS effect. Entries 5 and 6 refer to diketones. The delocalization of the anionic charge into two carbonyl oxygens in the respective enolate ions reduces the strength of hydrogen bonding solvation relative to that of nitronate ions. This reduces the solvational PNS effect as seen in the less dramatic solvent effect on the log ko values. The same is true for nitroacetophenone (entry 7) where the negative charge is shared between the nitro and the carbonyl group. Entries 8 and 9 involve cationic acids with the main resonance structure of the conjugate base being neutral, e.g., 10b. Hence hydrogen bonding solvation plays a minimal role and there

O– PhC

O +

CH

N 10a

CH3

PhC CH

N

CH3

10b

is no significant solvent dependence of ko. Entries 10–12 also show negligible solvent effects even though the conjugate bases of the respective carbon acids are anionic. In these cases the anionic charge is so highly dispersed that, once again, hydrogen bonding solvation is insignificant. More detailed analysis of the solvent effects on intrinsic rate constants which takes into consideration potential contributions from nonsynchronous solvation/desolvation of the carbon acid itself as well as of the proton acceptors and

242

C.F. BERNASCONI

their conjugate acids have been discussed in our previous review4 and elaborated upon in some subsequent studies.20,55 They involve the determination of solvent transfer activity coefficients and an estimate of the degree of solvational imbalance at the transition state.20,56,57 Even though a somewhat more refined picture emerges from such analysis, the broad qualitative conclusions stated above which focus on the solvation of the carbanions remains the same, at least with amines as the proton acceptor. For reactions involving carboxylate ion proton acceptors, the PNS effect of early desolvation of the carboxylate58 ion can make a significant contribution to the solvent effect on ko. This is illustrated by entries 4–6 of Table 4, which show a significantly larger change in log ko(RCOO) compared to log ko(R2NH). A complementary aspect of solvation is that it affects the magnitude of the transition state imbalance. This can be seen for the reactions of ArCH2NO2 in DMSO and MeCN where the imbalances are much smaller than in water (Table 2, entries 4 and 6). Again we see the connection between imbalance and intrinsic barriers: the greater imbalance induced by solvation leads to an enhanced intrinsic barrier. Polar effect of remote substituents One of the consequences of the imbalanced nature of the transition state is that the polar effect of a remote substituent may either increase or decrease the intrinsic barrier; whether there is an increase or decrease depends on the location of the substituent with respect to the site of charge development. Let us consider a reaction of the type shown in Equation (4). In this situation an electron-withdrawing substituent Z will decrease DG‡o or increase ko. This is because there is a disproportionately strong stabilization of the transition state compared to that of the product anion due to the closer proximity of Z to the charge at the transition state than in the anion. As discussed earlier, this also leads to an exalted Br½nsted aCH value and is the reason why aCH > bB for the deprotonation of carbon acids such as 11–13 and others (Table 2).

CH2NO2

CH2CH(COMe)CO2Et O Z

Z

Z 11

12

13

A different situation exists when the substituent is attached to the Y-group as schematically shown in Equation (14). In this case Z is closer to the negative charge in the anion than at

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

243

(14)

the transition state. Hence it is the carbanion rather than the transition state that receives a disproportionately strong stabilization by an electronwithdrawing substituent. The result is an increase in DG‡o or decrease in ko which translates into a depressed aCH value and is the reason why aCH < bB for the deprotonation of carbon acids such as 14–16 and others (Table 3). Note

that for 16, upon delocalization of the initially formed negative charge, there is neutralization of the positive charge on the pyridinium nitrogen to form a neutral conjugate base, as discussed earlier [Equation (7)]. Polar effect of adjacent substituents In principle, polar substituents directly attached to the carbon, Equation (15), should have a similar effect

(15)

on DG‡o or ko as in the situation described by Equation (4) for remote substituents, i.e., an electron-withdrawing substituent should reduce DG‡o or increase ko. In practice, it is difficult to quantify such effects because in most known examples factors other than the polar effect of Z contribute to changes in ko such as steric crowding, polarizability, hyperconjugation, and charge delocalization into Z. Nevertheless, there are several cases where changes in ko could definitely be attributed to the polar effect of Z. They are summarized in Table 519,20,25,42,43,54,55,66–70 which includes log ko values as well as Taft sF and sR values71 as measures of the polar and resonance effects, respectively, of Z.

244

Table 5 Effect of adjacent polar substituents on intrinsic rate constants Entry

C–H acid

1 H

sR

0.19

0.16

0.54

0.18

4.58b

log ko(RNH2)a

log ko(RCOO)a

References

2.84b

19

3.76b

19

CO2Me

2 H

log ko(R2NH)a

sF

CN

PhCOCH2

+ N CH3

0.29

0.16

2.29c

20

4

PhSO2CH2

+ N CH3

0.59

0.12

4.03c

20

5 6 7 8 9 10 11

CH3NO2 MeO2CCH2NO2 PhCOCH2NO2 CH3CH(NO2)2 HOCH2CH2NO2 PhCH2CH2NO2 PhSCH2NO2

0 0.19 0.29 0.64 0.13 0.04 0.29

0 0.16 0.16 0.16 –0.05 –0.08 –0.05

–0.5c 1.22d 1.57c

25 66 55 67 68 54 54

1.00c

d

–0.59 –1.16d 1.02d

–2.06d –0.13d

C.F. BERNASCONI

3

H

+ SMe2

13

F3CSO2CH2

14

F3CSO2CH2

SO2CF3

In units of M1 s1. In 50% DMSO–50% water (v/v) at 20C. c In water at 20C. d In water at 25C. e In 50% DMSO–50% water (v/v) at 25C. a

b

4.1d

1.01

0.13

42

0.83

0.26

5.0e

0.83

0.26

4.2e

69

70

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

12

245

246

C.F. BERNASCONI

Comparison of entry 2 with entry 1 shows a significant increase in ko that is caused by the greater electron-withdrawing strength of the CN (sF = 0.54) relative to that of the COOMe group (sF = 0.19). The CN group is also a somewhat stronger p-acceptor (sR = 0.18) than the COOMe (sR = 0.16) group which should slightly reduce ko and hence offset some of the increase resulting from the polar effect, i.e., the increase in ko due to the larger polar effect is probably somewhat greater than the observed increase in ko. On the other hand, the large size of the COOMe group could, in principle, lead to a steric reduction of ko for the COOMe derivative relative to the CN derivative. However, it was shown that this is not the case for these reactions which involve primary aliphatic amines as the proton acceptors.41 The change from PhCO (sF = 0.20) to the much more electron-withdrawing PhSO2 group (sF = 0.59) leads to a large increase in ko as seen from entries 3 and 4, respectively. In this case there may be a small contribution to the large difference in the ko value that arises from the stronger p-acceptor effect of the PhCO group (sR = 0.16) relative to that of the PhSO2 group (sR = 0.12) which reduces ko for the PhCO derivative relative to the PhSO2 derivative. In comparing entries 6 and 7 to entry 5 we note a substantial increase in ko when replacing a hydrogen with a COOMe or PhCO group. This implies that the main resonance structures of the corresponding anions are 17a and 18a, respectively, i.e., the COOMe and

CH3O

CH3O C

NO2–

CH

O

O

17a

CH

NO2

17b

O–

O PhC

C –

CH 18a

NO2–

PhC

CH

NO2

18b

PhCO groups act mainly through their polar effects. The somewhat larger ko for the PhCO derivative is consistent with the larger sF value of the PhCO group; the difference in the log ko values of 0.35 actually underestimates the true difference because ko for the PhCO derivative was determined at 20C rather than 25C. The higher intrinsic rate constant for 1,1-dinitroethane (entry 8) compared to that for CH3NO2 is open to two interpretations but both are related to the steric hindrance of the coplanarity of the two nitro groups in the anion. According to the first interpretation one nitro group in the anion is planar

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

247

and exerts its full p-acceptor effect while the second one is largely twisted out of the plane and enhances ko by its polar effect. The second, more likely, possibility is that both nitro groups are somewhat twisted out of plane so that neither exerts its full p-acceptor effect, while both exert their polar effect. For the entries 9–14 the potential polar effects of the respective substituents are either offset or enhanced by other factors such as hyperconjugation or polarizability as discussed in the next sections. Hyperconjugation Despite the fact that the HOCH2 group is mildly electron withdrawing (sF = 0.13), the log ko(R2NH) value for HOCH2CH2NO2 (Table 5, entry 9) does not show the expected increase relative to the respective log ko value for nitromethane (entry 5). This implies the presence of a ko-reducing factor that offsets the expected increase from the polar effect. In view of the higher temperature used for the reaction of HOCH2CH2NO2 (25C) compared to that with CH3NO2 (20C), the presence of a ko-reducing factor is even more compelling. A similar reduction in ko is seen in the deprotonation of PhCH2CH2NO2: even though the PhCH2 group exerts hardly any polar effect, log ko(R2NH) is measurably lower than for CH3NO2 (Table 5, entry 10). The reason for these reductions in ko has been identified as hyperconjugation in the anion (19b, R = OH or Ph).54,68

RCH2CH 19a

+ N

O– O–

H+ RCH

O– CH 19b

N O–

This hyperconjugation contributes to the stability of the respective nitronate ions as reflected in the reduction of the pKa values for HOCH2CH2NO2 (pKa = 9.40) and PhCH2CH2NO2 (pKa = 8.55) relative to CH3NO2 (pKa = 10.29). Inasmuch as hyperconjugation is expected to follow the same pattern as resonance/delocalization, i.e., being poorly developed at the transition state, its PNS effect should lower the intrinsic rate constant, as observed. Perhaps the best-known case of such hyperconjugation at work is seen when comparing pKa values and rate constants of the deprotonation of nitromethane, nitroethane, and 2-nitropropane by HO (Table 6). The pKa values show the increased hyperconjugative stabilization of the nitronate ion by one and two methyl groups, respectively (e.g., 19b, R = H). Since the transition state hardly benefits from hyperconjugation, the rate constants remain essentially unaffected by this factor and are mainly governed by the electron donating effect of the methyl groups which leads to a reduction in kOH.

248

C.F. BERNASCONI

Table 6 Deprotonation of nitroalkanes by HO in water at 25Ca C–H acid

pKCH a

kOH, M1 s1

CH3NO2 CH3CH2NO2 (CH3)2CHNO2

10.22 8.60 7.74

27.6 5.19 0.136

a

From Reference 51.

Interestingly, the fact that the trend in the rate constants is the opposite of that in the acidity constants translates into a negative aCH value of about –0.5 which has been called ‘‘nitroalkane anomaly.’’51 A somewhat similar situation has been observed in the deprotonation of the Fischer carbene complexes 20–24 summarized in Table 7.72,73 There is an increase in acidity in the order 20 < 21 < 22 and 23 < 24 which has been Table 7 Deprotonation of Fischer carbene complexes by HO in 50% MeCN–50% water (v/v) at 25C Entry

pKCH a

Carbene complex OMe

1

(CO)5Cr

C

2

(CO)5Cr

C

3

(CO)5Cr

C

4

(CO)5Cr

CH3 OMe CH2CH3 OMe CH(CH3)2

kaOH (M1 s1)

log ka,b o

(20)c

12.78

(21)c

12.62

23.4

0.09

(22)c

12.27

10.8

–0.99

(23)d

14.77

37.0

1.36

(24)d

13.41

39.5

0.70

152

1.07

O

O

5

(CO)5Cr CH3

a

Statistically corrected for the number of acidic protons.

b

Estimated based on the simplified Marcus equation log ko  0.5 log KOH with KOH = KCH a /Kw, Kw = 6.46  1016 M2. c

From Reference 72.

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

249

attributed to an increasing stabilization of the anions in the order 20 < 21 < 2272 and 23 < 2473 by the methyl groups. This effect reflects the well-

OMe (CO)5Cr

OMe

OMe

C

C

(CO)5Cr

(CO)5Cr

CH2

C C(CH3)2

CHCH3

20–

21–

22–

O

O (CO)5Cr

(CO)5Cr 23–

24– CH3

known stabilization of alkenes by methyl or alkyl groups74 and is commonly attributed to hyperconjugation as shown for the example of 21. The rate constants do not show the expected

OMe (CO)5Cr

OMe

C

(CO)5Cr CH

CH3

C CH

21–

CH2H+

21a–

increase with increasing acidity. This translates into a reduction of the intrinsic rate constants which is mainly due to the lag in charge delocalization into the (CO)5Cr group at the transition state (25). This lag not only prevents significant development of the hyperconjugative stabilization, there is also a destabilization of the transition state by the unfavorable interaction between the methyl group and the negative charge on the carbon. OMe (CO)5Cr

C

25

δ– CH

H

δ– OH

CH3

Polarizability The log ko(R2NH) value for PhSCH2NO2 of 1.02 (Table 5, entry 11) is about 1.6 log units greater than that for CH3NO2 (–0.59) and the pKCH of a

250

C.F. BERNASCONI

PhSCH2NO2 (6.67) is about 3.6 units lower than that of CH3NO2 (10.28). It was established that even though most of the increased acidity may be accounted for by the polar effect of the PhS group (sF = 0.29), at best one third or one half of the increase in log ko(R2NH) may be attributed to the polar effect.54 The rest was shown to be the result of a transition state stabilization of the negative charge on carbon by the polarizability effect of sulfur.54 The high intrinsic rate constant for the deprotonation of dimethyl-9fluorenylsulfonium ion (entry 12) has also been attributed to the polarizability of the sulfur atom.42 The same is true for the effect of the trifluoromethylsulfonyl group on the benzyltriflones in entries 13 and 14. The latter group has a very high sF value (0.83) and hence its electron-withdrawing inductive effect certainly contributes significantly to the very high ko values. However, its sR value is also very high (0.26) which implies that the resonance effect may offset a significant fraction of the ko-enhancing inductive effect. Hence, the high ko values strongly suggest a major contribution by the polarizability effect. This conclusion is strongly supported by 1H, 13C, and 19F NMR data as well as solvent effect studies on the kinetic and thermodynamic acidities of various trifluoromethylsulfonyl derivatives from Terrier’s laboratory.69,70,75,76 There is a broader significance to the conclusion regarding the importance of polarizability effects on intrinsic barriers because it also addresses the question whether d–p p bonding or negative hyperconjugation in the anion may play a significant role. The notion that d–p bonding between the carbanion lone pair and the sulfur 3d orbital (e.g., 26) may account for the stabilization of carbanions was promoted by numerous authors,77–82 although

CH3

S

CH –2

CH3

26a

S

CH2

26b

theoretical work later challenged this idea, suggesting the polarizability of sulfur as the main source of stabilization.83–86 A third potential interaction mechanism, negative hyperconjugation (e.g., 27), has also been invoked by several authors,84,86–88 and Wiberg et al.89 as well as Cuevas and Juaristi90 have concluded that this may be the main factor in the stabilization of the dimethylsulfide ion by sulfur.

CH3

S 27a

CH –2

CH3

S

CH 2

27b

If d–p bonding or negative hyperconjugation were to play an important role in the stabilization of the conjugate bases of the dimethyl-9flurorenylsulfonium ion, the benzyltriflones, or thiophenylnitromethane, one

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

251

would expect a reduction in the intrinsic rate constant. This is because these interaction mechanisms belong to the broad category of resonance effects and hence their development at the transition state is expected to lag behind proton transfer and lower the intrinsic rate constant. The large increase in log ko(R2NH) for the deprotonation of PhSCH2NO2 relative to that for CH3NO2 and the unusually high ko values in entries 12–14 indicate that the influence of these factors on ko is negligible compared to the polarizability effect. A further comment regarding the nature of the ko-enhancing effect by a polarizable substituent is in order. It is quite similar to that exerted by the polar effect of an electron-withdrawing group in that it leads to a disproportionately strong stabilization of the transition state due to the closer proximity of the substituent to the site of the incipient negative charge. However, there are some important differences between the two effects that result from the fact that polarizability effects are proportional to the square of the charge91 and drop off very steeply with the 4th power of distance,91 whereas polar effects are simply proportional to the charge and drop off only with the square of the distance.91 The steep drop off with distance may potentially render polarizability effects on intrinsic rate constants more dramatic than polar effects but only when the imbalance is large and proton transfer has made significant progress at the transition state. A more detailed discussion of these points has been presented elsewhere.3,41 Electrostatic effects Electrostatic effects may significantly affect intrinsic barriers or intrinsic rate constants, especially when there is a positive charge directly adjacent to the carbon that gets deprotonated, as exemplified by Equation (16). Keeffe and Kresge92 have shown that a large body of data on the ‡ + N

CH2 28

C

O–

O

O Ph

+ N

δ– CH

C

Ph

+ N

CH

C

Ph + H2O

(16)

H δ– OH

deprotonation of simple aldehydes and ketones by HO in water obey a linear correlation between log(kOH/p) and log ðKCH a =qÞ over a range of 11 pKa units. The points on the correlation may be understood to refer to reactions with equal or at least comparable intrinsic rate constants, while deviating points indicate higher or lower intrinsic rate constants. The point for the reaction of Equation (16)93 shows a positive deviation of almost 3 log units92 which suggests a strongly enhanced intrinsic rate constant. This increase in ko can

252

C.F. BERNASCONI

be attributed to a combination of the PNS effect by the polar effect of the pyridinio group and the electrostatic stabilization of the negative charge at the transition state which is disproportionately strong compared to that of the enolate ion due to its closer proximity to the positive charge. The deprotonation of 29 by HO93 also shows a positive deviation from the Keeffe/Kresge plot but it amounts to only about 1 log unit, as

CH2

Ph C O

N + CH3 29

expected due to the greater distance of the positive charge from the reaction site. In fact, it is not clear whether in this case the entire acceleration might be due to a polar rather than an electrostatic effect. An interesting case is the deprotonation of 30 by HO94 with a rate constant, if placed on the Keeffe/Kresge plot, would deviate positively by about 0.34 log units. The following

‡

HO– +

O

N + CH3

C CH3

N + CH3

O C δ– CH2

N

+ CH3

C

CH2 + H2O

O–

H

30

31

δ– OH

32

interpretation was given. The likely conformation of the transition state is one with the closest distance between the charges as shown in 31, while the same is true for the product 32 except that at the transition state the negative charge mainly resides on the carbon while in the product ion (32) it mainly resides on the oxygen. Hence, even though the distances between the charges are about the same in 31 and 32, to the extent that carbon is much less able to support a negative charge than oxygen, the transition state derives a disproportionately large degree of stabilization from the electrostatic effect of the positive charge compared to the product; the stabilization of the latter by the positive charge is further attenuated by the strong solvation of the anionic oxygen.

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

253

The above interpretation is supported by data on the deprotonation of 33 by HO.95 In this situation, it is only the product 35 that benefits from electrostatic stabilization while the ‡ HO– +

+ H2O

δ–

N + CH3

O

N + CH3

H O

33

N + CH3

δ– OH

34

O–

35

negative charge at the transition state 34 is too far for a significant interaction with the positive nitrogen. In the terminology of the PNS this is a case where the product stabilization factor lags behind proton transfer and hence ko should be reduced. This is borne out by the fact that when the rate constant of this reaction is included on the Keeffe/Kresge correlation it shows a negative deviation of 0.85 log units. Whether the entire 0.85 log units deviation should be attributed to this PNS effect is not clear because cyclohexanone also shows a slight negative deviation from the Br½nsted line. p-Donor effects Carbon acids activated by strong p-acceptors that also contain a p-donor capable of interacting with the p-acceptor in a push–pull fashion pose an interesting problem. Fischer carbene complexes such as 20,94 23,97 36,96 37,96 38,98 and 3998 fall into this category. Thermodynamic

OCH3 (CO)5Cr

C CH3 20

OCH3 (CO)5W

C CH3

C

37

SCH3 (CO)5Cr

C CH3

23

C CH3

36

O (CO)5Cr

OCH2CH3 (CO)5Cr

38

SCH3 (CO)5W

C CH3 39

and kinetic data for these carbene complexes are reported in Table 8.96–98 There is a strong correlation between the strength of the p-donor and the pKa

254

Table 8 Acidities of Fischer carbene complexes and intrinsic rate constants for their deprotonation by secondary alicyclic amines, primary aliphatic amines and hydroxide ion in 50% MeCN–50% water (v/v) at 25C pKa

log ko(R2NH)

log ko(RNH2)

log ko(HO)

References

(36)

12.36

3.18

2.72

1.09

96

(20)

12.50

3.70

3.04

1.31

96

(37)

12.98

1.38

96

(23)

14.47

1.51

97

(39)

8.37

2.51

2.50

98

(38)

9.05

2.61

2.09

98

Entry

C

1

(CO)5W

2

(CO)5Cr

C

3

(CO)5Cr

C

4

(CO)5Cr

5

(CO)5W

6

(CO)5Cr

OCH3 CH3 OMe CH3 OCH2CH3 CH3 O

C

SCH3

C

SCH3 CH3

C.F. BERNASCONI

CH3

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

255

values, i.e., the stronger the p-donor, the lower the acidity of the carbene complex. For example, the pKa of 36 is about 4.0 units higher than that of 39 while that of 20 is about 3.5 units higher than that of 38. This reflects the fact that the MeO group is a stronger p-donor (sR = –0.43)71 than the MeS group (sR = –0.15)71 and leads to stronger resonance stabilization (40, 40) of the carbene complex which reduces its

+ XCH3

XCH3 (CO)5M

C 40

(CO)5M CH3

40±

M = Cr or W X = O or S

C CH3

acidity. The cyclic complex 23 is of particular interest with a pKa that is 2 units higher than that of 20. In this case the resonance stabilization of the carbene complex is enhanced by virtue of its cyclic structure in which the oxygen atom is locked into a position for better p-overlap with the carbene carbon.97 53Cr NMR data are in agreement with this interpretation.99 An even greater reduction in acidity is observed for the Me2N derivative, 41 (sR = –0.56)71

NMe2 (CO)5Cr

C CH3 41

which makes it impossible to determine its pKCH in an aqueous environment. a However, a pKCH  32.5 was estimated in acetonitrile which is more than 10 a pKa units higher than the pKCH 22.2 of 20 in the same solvent.100 a The p-donor effect on intrinsic rate constants is more difficult to interpret. For the alkoxy carbene complexes the log ko(HO) values are all about the same within the experimental uncertainty, suggesting that changing p-donor strength has no significant effect on the intrinsic barriers. This seems surprising because one might have expected that the push–pull resonance would follow the same PNS rules as resonance delocalization, i.e., it should lower the intrinsic rate constants due to early loss of the resonance effect at the transition state. The most plausible explanation for the results is that there is another factor that offsets the ko-lowering PNS effect. It has been suggested that this other factor comes from an attenuation of the lag in the carbanionic resonance development by the p-acceptor because the contribution of 40 to the structure of the carbene complex leads to a preorganization of the (CO)5M-moiety

256

C.F. BERNASCONI

toward its electronic configuration in the anion. This preorganization is likely to stabilize the transition state by facilitating the delocalization of the negative charge into the (CO)5M-moiety, i.e., by reducing the degree of imbalance.96–98 Additionally or alternatively, the electrostatic interaction between the partial positive charge on X and the partial negative charge at the carbon of the transition state (42) is expected to stabilize the latter and lower the intrinsic barrier. δ+ XCH3

δ– (CO)5Cr

C

42

δ– CH2

H

Bν + δ

The reduction in the imbalance by preorganizing the carbene complex structure in the manner described above also manifests itself in the Br½nsted aCH and bB values for the deprotonation of phenyl-substituted (benzylmethoxycarbene) pentacarbonyl chromium (43-Z) by amines.101

OCH3 (CO)5Cr

C

δ+ OCH3

δ– Z

(CO)5Cr

C

CH2

δ– Z

(CO)5Cr

CH2

43-Z

43-Z′

C

δ+ OCH3 δ– CH

Z

44-Z H B

δ+

The results are summarized in Table 9. They show that, within experimental error, aCH  bB rather than the expected aCH > bB. It appears, then, that the p-donor effect of the methoxy group masks the true extent of the imbalance by reducing aCH. One way to understand this reduction is to assume that the partial positive charge on the MeO group of the resonance hybrid 43-Z0 is largely maintained at the transition state 44-Z. This means that the stabilizing effect of an electron-withdrawing Z-substituent on the negatively charged carbon at the

Table 9 Reactions of 43-Z with amines in 50% MeCN–50% water (v/v) at 25Ca Reaction 43-H þ R2NH 43-H þ RNH2 a

From Reference 101.

bB

Reaction

aCH

0.48  0.07 0.54  0.04

43-Z þ piperidine 43-Z þ n-BuNH2

0.530.02 0.560.03

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

257

Table 10 Average aCH and bB values for the deprotonation of 45-Z and 46-Z by amines in 50% MeCN–50% water (v/v) at 25Ca Parameter

R2NH 45-Z 0.41 0.55 –0.15 46-Z 0.33 0.45 –0.12

aCH bB aCH – bB aCH bB aCH – bB a

RNH2 0.37 0.73 –0.36 0.33 0.47 –0.14

From Reference 102.

transition state is partially offset by its destabilizing effect on the positively charged oxygen and hence aCH is reduced.102 Supporting evidence for the above explanation comes from the study of the deprotonation of 45-Z and 46-Z by a series of amines.102 Average aCH and bB values are summarized in Table 10.

Z (CO)5Cr

Z

OCH2

S

C

(CO)5Cr

CH3

C CH3

45-Z

46-Z

Taking 45-Z as an example, the reaction can be described by Equation (17). The observation ‡ δ+ OCH2

δ– B + (CO)5W

C CH3

Z

δ– (CO)5W

C

δ+ OCH2 δ– CH2

Z

H

B δ+

45-Z′

(17) Z OCH2 (CO)5W

C

+ BH+

CH2

that aCH < bB is reminiscent of the situation described in Equation (6) and other cases summarized in Table 3. In both situations there is creation of partial negative charge on the carbon at the transition state and in both cases the Z-substituent is located far away from that carbon. In the

258

C.F. BERNASCONI

deprotonation of 2-nitro-4-Z-phenylacetonitrile [Equation (6)], the negative charge moves closer to Z and becomes a full charge in the product ion; in the deprotonation of 45-Z the negative charge also moves closer to Z but here the effect is the neutralization of positive charge on the oxygen. In terms of substituent effects, i.e., aCH versus bB, the outcome is the same in both situations. A further example of an imbalance reduction is seen in the reaction 1benzyl-1-methoxy-2-nitroethylenes (47-Z) with secondary alicyclic amines in 50% DMSO–50% water (v/v).103

OMe

H C

C

O2N

Z CH2

47-Z

47-Z is analogous to 43-Z. The reported Br½nsted coefficients are aCH = 0.84 and bB = 0.47. These values still show a large imbalance (aCH – bB = 0.37) but it is much smaller than that for the deprotonation of phenylnitromethanes (aCH = 1.29, bB = 0.56, aCH – bB = 0.73). Turning to the effect of the heteroatom, the change from MeX = MeO to MeS (36 vs. 39 and 20 vs. 38 in Table 8) leads to a decrease in log ko(R2NH) of 0.67 (M = W) and of 1.09 (M = Cr) log units, respectively, and a decrease in log ko(RNH2) of 0.22 (M = W) and of 0.95 (M = Cr) log units, respectively. Even though weaker, due to the reduced p-donor strength of the MeS group, the two compensating factors discussed for the alkoxy derivatives may still essentially offset each other. This means that the decreases in ko for the sulfur derivatives must be due to other factors. One such factor is early developing steric crowding at the transition state due to the larger size of the sulfur atom. Another is the weaker polar effect of the MeS compared to that of the MeO group which favors the methoxy derivative. Aromaticity The question how aromaticity in a reactant or product might affect intrinsic barriers has only recently received serious attention. Inasmuch as aromaticity is related to resonance one might expect that its development at the transition state should also lag behind proton transfer (or its loss from a reactant would be ahead of proton transfer) and hence lead to an increase in DG‡o , as is the case for resonance/delocalized systems. However, recent studies from our laboratory suggest the opposite behavior. The first such study involved the deprotonation of the cationic rhenium Fischer carbene complexes 48Hþ-X by primary aliphatic amines, secondary

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

259

alicyclic amines, and carboxylate ions in aqueous acetonitrile.104 The conjugate bases, 48-X, represent heterocyclic aromatic X Bν

+ + Cp(NO)(PPh3)Re

X

k1 k–1

Cp(NO)(PPh3)Re

48H+-O

+ BHν + 1

48-O (X = O) 48-Se (X = Se) 48-S (X = S)

(X = O) 48H+-Se (X = Se) 48H+-S (X = S)

(18) derivatives of furan, selenophene, and thiophene, respectively. The increase in aromaticity along the order 48-O < 48-Se < 48-S105 is reflected in the decreasing pKa values of the respective 48Hþ-X (Table 11). The intrinsic rate constants for proton transfer follow the order ko(X = O) < ko(X = Se) < ko(X = S) for each buffer family (Table 11), i.e., they increase with increasing aromaticity of 48-X. According to the PNS, these results imply that development of the aromatic stabilization of 48-X has made more progress at the transition state than proton transfer. A second system showing similar results is that of Equation (19).108 Table 12 summarizes pKCH values and intrinsic rate constants for the reactions with a primary aliphatic and secondary O–

O k1

Bν + X

+ BHν + 1

k–1

(19)

X 49–-O (X = O) 49–-S (X = S)

49H-O (X = O) 49H-S (X = S)

alicyclic amines in aqueous solution. Again, the stronger aromatic stabilization of the sulfur heterocyclic system is reflected in the greater acidity of 49H-S Table 11 Acidities of rhenium Fischer carbene complexes and intrinsic rate constants for their deprotonation by amines and carboxylate ions in 50% MeCN–50% water (v/v) at 25Ca,b Carbene complex

pKCH a

log ko(R2NH)

log ko(RNH2)

log ko(RCOO)

48Hþ-O 48Hþ-Se 48Hþ-S

5.78 4.18 2.50

–0.46 0.92 1.05

–0.88 0.14 0.27

–0.01 0.72 1.21

a b

ko in units of M1 s1. From Reference 104.

260

C.F. BERNASCONI

Table 12 Acidities of benzofuranone and banzothiophenone and intrinsic rate constants for their deprotonation by amines in water at 25Ca,b C–H Acid

pKCH a

log ko(R2NH)

log ko(RNH2)

49H-O 49H-S

11.72 9.45

1.64 2.64

1.16 1.72

a b

ko in units of M1 s1. From Reference 108.

(pKa = 9.45) compared to that of 49H-O (pKa = 11.72) and again the intrinsic rate constants are greater for the more aromatic system. A third reaction, Equation (20), yielded somewhat ambiguous results stemming from k1

Bν +

O– + BHν + 1

O k–1

X

(20)

X 50–-O

50H-O (X = O) 50H-S (X = S)

(X = O) 50–-S (X = S)

complications due to the p-donor effects of the ring heteroatom (50H-X).109 However, a

O X 50H-X

O–

+ X ± 50H-X

detailed analysis108 revealed that here again the greater aromaticity of the sulfur derivative increases ko relative to that for the oxygen derivative. An interesting situation exists in the deprotonation of the pentacarbonyl(cyclobutenylidene)chromium complexes 51 and 52.110 These complexes are characterized by a strong Me

Ph

(CO)5Cr

NEt2 Me

H 51

(CO)5Cr

NEt2 Me

H 52

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

261

push–pull interaction represented by 53a/53b but their most interesting feature is that their R

R

(CO)5Cr Me

H

+ NEt2

(CO)5Cr

NEt2

Me

R = Me or Ph

53a

H 53b

conjugate bases are derivatives of cyclobutadiene (54) which makes them antiaromatic. This anti-aromaticity is reflected in the very high pKa values determined in acetonitrile and the R Me2N (CO)5Cr

NEt2 Me

54 (R = Me or Ph)

Me2N

P

N

Me2N

NMe2 + NEt P NMe2

P2-Et

extremely low calculated gas phase acidities.110 The intrinsic rate constants for the reaction of 51 with the phosphazene base P2-Et111 in acetonitrile suggests that the anti-aromaticity of the anion has an intrinsic barrier-lowering effect, although this conclusion was tentative because ko for this reaction is affected by several other factors. The implication of this result is that the development of anti-aromaticity at the transition state may lag behind proton transfer. Possible reasons why transition state aromaticity is able to develop early while resonance development lags behind proton transfer at the transition state, and why anti-aromaticity lags behind proton transfer, will be discussed in the section on ab initio calculations. These calculations have provided important additional insights because they allow a direct probe of transition state aromaticity or anti-aromaticity.

3

Proton transfers in the gas phase: ab initio calculations

THE CH3Y/CH2=Y SYSTEMS

The computational investigation of identity proton transfers such as Equation (21) in the gas phase has been particularly useful because the barriers of such reactions are the intrinsic barriers and the

262 Y

C.F. BERNASCONI CH3 + CH2

Y–

–Y

CH2 + CH3

(21)

Y

transition state is symmetrical with respect to the proton transfer (50%) which makes it easy to recognize the presence of imbalances. These studies have provided further insights into the PNS that complement those gained from solution phase reactions. The p-acceptor groups Y examined include CH=O,112–116,118 NO2,117,118 NO,118 CH=CH2,114,115,118,119 C N,114,118 þ

þ

CH=NH,118 CH=S,118 CH=OH ,120 and NO2 H.117 The most comprehensive study which also incorporates results form earlier work is that by Bernasconi and Wenzel118; the present discussion is largely based on this paper and on references 113, 117, and 120. A major conclusion is that even though the intrinsic barriers of these gas phase reactions depend on the same factors as solution phase proton transfers such as resonance, polar, and polarizability effects, the relative importance of these factors is quite different in the gas phase, and electrostatic effects involving the proton-in-flight constitute an important additional factor. Evidence of imbalance One of the first questions we113,117,118 and others112,114,117 asked is whether the Y-group induces similar transition state imbalances as observed in solution. Calculation of geometric parameters such as bond lengths and angles as well as group charges indicate that the transition states of these reactions are indeed imbalanced in the sense that charge delocalization into the Y-group lags behind proton transfer. In order to quantify the degree of charge imbalance for a given reaction we developed a formalism based on Equation (22)113; note that in Equation (22) B stands for

–1 + χ – χ C Y

and the full representation of the

transition state is given by 55. Equation (22) represents a further ‡ –1 + δB δH

B– + H

C

Y

B

H

–δC –δY

C

–1 + χ –χ Y + BH C

Y

(22)

refinement of Equation (9) which was a more nuanced version of Equation (3); it not only allows for a certain

–δY –δC

δH

Y

H

C

55

–δC –δY

C

Y

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

263

Table 13 Relative contraction of C–Y bond in the anion and negative charge on Y-group of aniona in the CH3Y/CH2=Y and NCCH2Y/NCCH=Y systems CH3Y/CH2=Yb Y

100  DroCY =rC  Y d

CN C CH CH=CH2 CH=NH(syn) CH=NH(anti) CH=O CH=S NO2 NO þ CH=OHf,g þ NO2 Hf,h

4.10 6.69 6.86 7.11 7.14 7.64 8.94 9.18 10.8 8.26 11.3

Charge on Ye –0.356 –0.587 –0.539 –0.507 –0.548 –0.531 –0.756 –0.854 –0.866 0.086 –0.171

NCCH2Y/NCCH=Yc 100  DroCY =rC  Y d

Charge on Ye

4.63

–0.232

6.16

–0.312

8.64 8.78 10.4 11.9

–0.343 –0.544 –0.654 –0.683

a

MP2/6-311þG(d,p). From Reference 118. c From Reference 140. d % reduction of C–Y bond length upon conversion of CH3Y into CH2=Y. e NPA at MP2/MP2. f Here the conjugate base is neutral. g From Reference 117. h From Reference 114. b

amount of charge delocalization at the transition state (–d Y, Equation (9)) but also takes into account that the negative charge in the anion is not necessarily completely delocalized into the Y-group (–w) and that some positive charge may develop on the proton-in-flight at the transition state (dH). The results summarized in Table 13 indicate that even for the strongest pacceptors, NO2 and NO, the charge on Y in the anion, –w, is somewhat less than –1.0 and substantially less than –1.0 for the weaker p-acceptors. Furthermore, the results reported in Table 14 show that there is in fact a substantial positive charge on the proton-in-flight in the range of 0.3 in all cases. The refinements introduced into Equation (22) require corresponding modifications to Equations (10–12), i.e., Equation (10) becomes Equation (23), Equation (11) becomes Equation (24), and Equation (12) becomes Equation (25). The negative charge d Y  pb:o: ðd B þ dY Þu

ð23Þ

264

Table 14 NPA group charges in the CH3Y/CH2=Y systemsa,b CH3

CH2=Y

Differencec

CH3C N CH3(CH2) C N H (transferred)

0.041 –0.041

–0.644 –0.358

–0.685 –0.315

–0.446 –0.208 0.303

–0.487 –0.165

CH3C CH CH3(CH2) C CH H (transferred)

0.028 –0.028

–0.463 –0.537

–0.491 –0.509

–0.429 –0.219 0.296

–0.457 –0.191

CH3CH=CH2 CH3(CH2) CH=CH2 H (transferred)

0.003 –0.003

–0.461 –0.539

–0.464 –0.536

–0.376 –0.266 0.285

–0.379 –0.263

CH3CH=NH (syn) CH3(CH2) CH=NH H (transferred)

–0.013 0.013

–0.493 –0.507

–0.480 –0.520

–0.398 –0.246 0.293

–0.386 –0.261

CH3CH=NH (anti) CH3(CH2) CH=NH H (transferred)

0.004 –0.004

–0.452 –0.548

–0.456 –0.544

–0.366 –0.281 0.293

–0.370 –0.277

CH3CH=O CH3(CH2) CH=O H (transferred)

–0.021 0.021

–0.469 –0.531

–0.448 –0.522

–0.384 –0.266 0.301

–0.363 –0.287

Group

TS

Differenced

C.F. BERNASCONI

0.021 –0.021

–0.244 –0.756

–0.265 –0.785

–0.233 –0.413 0.291

–0.254 –0.392

CH3NO2 CH3(CH2) NO2 H (transferred)

0.244 –0.244

–0.146 –0.854

–0.390 –0.610

–0.093 (–0.060)g –0.535 (–0.582)g 0.253 (0.286)g

–0.337 (–0.304)g –0.291 (–0.388)g

CH3NO CH3(CH2) NO H (transferred)

0.155 –0.155

–0.134 –0.866

–0.289 –0.711

–0.082 –0.548 0.260

–0.237 –0.393

CH3CH=OHþe CH3(CH2) CH=OHþ H (transferred)

0.149 0.851

–0.086 0.086

–0.235 –0.765

–0.132 0.472 0.320

–0.281 –0.379

CH3NO2Hþf CH3(CH2) NO2Hþ H (transferred)

0.403 0.597

0.171 –0.171

–0.232 –0.768

0.172 0.207 0.240

–0.231 –0.390

a

MP2/6-311þG(d,p).

b

From Reference 118.

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

CH3CH=S CH3(CH2) CH=S H (transferred)

c

Difference between anion and neutral. On the Y-group this difference corresponds to w in Equations (12–14). d Difference between TS and neutral; it corresponds to –d C and –d Y in Equations (12–14), respectively. From Reference 120.

f

From Reference 117.

g

CH3 NOþ 2 H/CH2=NO2H system, from Reference 117.

265

e

266

C.F. BERNASCONI

pb:o:  wðd C þ dY Þ

ð24Þ

d Y  wðd C þ d B Þn

ð25Þ

transferred from the base to the CH2Y fragment of the transition state is now d B þ dH which is equal to d C þ dY, while the p-bond order is not only related to the transferred charge but also to the degree of charge delocalization in the anion (w). Furthermore, there is no requirement that the exponents u and v be exactly 1.0 or that n be exactly 2.0; as pointed out by Kresge,51 there may not be a strict proportionality between dY and (dC þ d Y) and hence n = u þ v may be £2 or 2. Note, however, that imbalance in the sense of delayed charge delocalization requires n > 1; this is because such an imbalance implies that the ratio of the charge on Y to the charge on C is smaller at the transition state than in the anion, i.e., dC/dY < w/1– w and this is only possible for n > 1. By the same token, n = 1 would mean that delocalization is synchronous with proton transfer. Note also that Equations (23–25) are not only valid at the transition state but at any point along the reaction coordinate, including the final products where d C þ dY = 1 and hence dY = w. Table 15 summarizes the imbalance parameters n calculated from Equation (26) which is the logarithmic version of Equation (25) solved for n. The n values range from 1.28 to 1.61 in most cases

n¼

log ðdY =wÞ log ðd C þ d Y Þ

ð26Þ

except for CH3C CH where n = 2.26.121 These numbers suggest that in the gas phase the imbalances are relatively small and substantially smaller than those estimated for proton transfers in hydroxylic solvents.3,4 Calculations by Yamataka et al. of the deprotonation of several nitroalkanes by CN122 and 123 HOðH2 OÞ also indicate sharply reduced imbalances. For example, the 2 Br½nsted aCH value for the deprotonation of substituted phenylnitromethanes is ‘‘normal,’’ i.e., < 1; this contrasts with aCH = 1.54 or 1.29 for the deprotonation of substituted phenylnitromethanes in water by HO or piperidine, respectively (Table 2). Furthermore, aCH based on a comparison between CH3NO2 and CH3CH2NO2 is also normal, i.e., aCH > 0122 rather than < 0 (‘‘nitroalkane anomaly,’’ Table 6). These results are not all that surprising in view of the absence of the strong solvational contribution to the imbalance in hydroxylic solvents discussed earlier. Table 15 includes two geometric parameters that provide complementary information about transition state imbalances. They are the % progress of the

CH3Y/CH2=Ya

Y

CN C CH CH=CH2 CH=NH(syn) CH=NH(anti) CH=O CH=O(constr)f CH=S NO2 NO2 (constr)f NO CH=OHþg h NOþ 2H a

NCCH2Y/NCCH=Yb

nc

100  ðDr‡CY =DroCY Þd

100  (Da‡/Dao)e

nc

100  (Dr‡CY =DroCY )d

100  (Da‡/Dao)e

1.51 2.26 1.61 1.58 1.55 1.52 1.10 1.42 1.59 1.33 1.28 1.69 1.42

53.3 34.7 56.3 57.0 61.7 65.2 73.8 64.2 57.7 71.5 70.0 62.8 64.9

21.2 9.0 22.6 25.8 27.6 34.1 100 41.0 26.8 100 44.0 27.6 27.2

1.94

44.1

11.1

2.14

48.4

20.0

1.99

60.2

33.0

1.85 2.06

60.9 56.9

37.7 23.1

1.67

66.1

29.4

From Reference 118. From Reference 140. c From Equation (26). d % Progress of C–Y bond contraction. e % Change in pyramidal angle. f Transition state geometry constrained to be planar with a = 0. g From Reference 120. h From Reference 117.

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

Table 15 Imbalance parameter n, progress in C–Y bond contraction and planarization of the a-carbon at the transition state in the CH3Y/CH2=Y and NCCH2Y/NCCH=Y systems

b

267

268

C.F. BERNASCONI

C–Y bond contractions and of the planarization of the a-carbon as measured by the % progress in the change of the pyramidal angle a. The pyramidal angle is defined as shown in 56 where the dashed lines are the projection

C

Y

H H

α

56

of the C–Y bond and the bisector of the HCH group, respectively; note that for a planar molecule or ion (sp2-carbon) a is zero. There is a strong inverse correlation between n and the progress in the planarization as well as the progress in the C–Y bond contraction. The relatively small Da‡/Dao values indicate substantial retention of the sp3 character of the a-carbon that is particularly pronounced for CH3C CH which has the largest n value, but still appreciable for CH3NO and CH3CH=S which have the smallest n values. þ Tables 13–15 include results for the CH3CH=OH/CH2=CHOH and þ

CH3 NO2 H/CH2=NO2H systems; the patterns relating to the C–Y bond contraction, planarization, and charge changes are quite similar to those observed for the respective neutral/anion systems. Regarding the imbalance parameter n, it is still given by Equation (26) because, even though the absolute charges on the various sites are different as shown in the reaction scheme of Equation (27), it is the charge changes during the reaction that determine the imbalance. ‡

B+H

C

+ YH

δB

δH

B

H

–δC 1–δY C

YH

–1 + χ 1 – χ YH + BH+ C

(27)

Additional insights into the reasons for the presence of transition state imbalances have come from the study of the CH3CH=O/CH2=CH–O113 117 and CH3NO2/CH2=NO systems for which the geometries of the transition 2 states were constrained to be planar, i.e., a = 0. The results are included in Table 15. The consequence of these constraints is a reduced n value and greater progress in the C–Y bond contraction. However, as will be discussed later, this greater charge delocalization does not result in a lower barrier but rather in a higher barrier. Similar conclusions were reached by Lee et al.116 and by the Saunders–Shaik group115,119 as discussed in more detail below.

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

269

More O’Ferrall–Jencks diagrams Reactions with imbalanced transition states are conveniently described by More O’Ferrall124–Jencks125 diagrams. Such a diagram is shown in Fig. 1 for the deprotonation of a nitroalkane by the general base B which, e.g., could be an amine or HO. In reactions where both the proton donor and acceptor contain the p-acceptor group Y as in Equation (21), there is a two-fold imbalance, one for the lag in the charge delocalization into the p-acceptor of the incipient product ion and one for the advanced localization of the charge from the reactant anion. This two-fold imbalance can be represented by a sixcorner diagram; such a diagram is shown in Fig. 2 for the CH3NO2/CH2= 117 NO Corners 1 and 4 are the reactants and products, respectively. 2 system. Corners 2 and 3 are hypothetical states where the respective nitromethide ions have their charge localized on the a-carbon (57), while corners 5 and 6 are hypothetical states where the nitromethane is polarized in the manner shown in 58.

– CH2

NO2

+ HCH2

57

RCH

H+-transfer

NO–2 + BHν + 1

Delocalization

NO–2

58

Delocalization

H+ Bν + RCH

– NO2

‡

Bν + RCH2NO2

H+-transfer

–

RCHNO2 + BHν + 1

Fig. 1 More O’Ferrall–Jencks diagram for the deprotonation of a nitroalkane. The curved line shows the reaction coordinate with charge delocalization lagging behind proton transfer.

270

C.F. BERNASCONI NO– 2

CH2 CH3

NO2 4C ) t ha c rg du

ne

pro

es

hif

a eth

t (n

m itro

+ HCH2 CH2

ion

6 – NO2 C ha rge NO –

sh

ift

pro

du

g

ar – NO2 Ch 5

2

te

(ni

ate

tro

me

tha

ne

cta

nt)

CH3 CH2

s rge

NO2

2 3 nt) CH – a t c CH2 rea

NO2 NO2

n itro

t (n

hif

rea

ion

NO2

3

TS (optim)

• •

– ct) CH 2 3 CH

Proton transfer

es

TS (constr)

2

na

hi

Proton transfer

CH2 + HCH2

NO–

itro

n ft (

a

1

Ch

NO2

NO2–

Fig. 2 Modified More O’Ferrall–Jencks diagram for the CH3NO2/CH2=NO 2 system. The curved lines represent the reaction coordinates through the optimized and constrained transition state, respectively. The constrained transition state is less imbalanced as indicated by its location to the left of the optimized transition state.

Calculations suggest that corners 5 and 6 are 34.3 kcal mol1 above the level of the reactant/product corners while corners 2 and 3 are about 9.4 kcal mol1 above the reactant/product corners.117 This indicates that the energy surface defined by the diagram exhibits a strong downward tilt from left to right, suggesting that the reaction coordinate and transition state should be located in the right half of the diagram. This is consistent with the observed imbalance according to which charge shift from the nitro group toward the carbon of the reactant nitromethide anion is ahead of proton transfer and the charge shift from the carbon to the nitro group in the incipient product nitromethide anion

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

271

lags behind proton transfer. We also note that the smaller imbalance observed for the reaction through the geometrically constrained (a = 0) transition state (Table 15) requires placement of this transition state to the left of the optimized transition state, but still in the right half of the diagram; this move to the left is also in the direction of increased energy, consistent with the higher energy calculated for the constrained transition state (see below). The features and conclusions from Fig. 2 are very similar to those obtained from a similar diagram for the CH3–CH=O/CH2=CHO system,113 including the left-to-right downward tilt of the surface and the relative placement of TS(optim) and TS(constr) within the right half of the diagram. Valence bond analysis of the imbalance The Shaik–Pross valence bond (VB) configuration-mixing model mentioned earlier has been further developed and applied to Equation (21) with Y = CH=CH2115,119 and CH=O.115 Specifically, the relative importance of the various contributing VB structures to the optimized transition state structure was determined by the self-consistent field valence bond (VBSCF) method and the effect on the energy evaluated. The calculated VB structures are shown in Chart 2. The numbers in parentheses are the relative weights contributed to the transition state structure by each VB structure: the first numbers refer to X = CH2, the second to X = O. These weights were calculated by the ‘‘localized valence bond method’’ (LVB)115; the authors also calculated weights by the

X

CH

CH2

– H :CH2

CH

X

X

– CH CH2: H

X

CH

CH2

H CH2

CH2

CH

X

φ3 (0.187, 0.151)

φ1 (0.187, 0.151)

CH

X:–

–:X

CH

CH2 H

CH2

CH

X

CH

X

φ4 (0.049, 0.118)

φ2 (0.049, 0.118)

X

CH

+ – – CH2: H :CH2

CH

X

φ5 (0.402, 0.540) X

CH

+ – CH2: H CH2

CH

X:–

–:X

CH

φ6 (0.059, 0.102)

φ7 (0.059, 0.102) –:X

CH

+ CH2 H CH2

CH

φ8 (0.006, 0.029)

Chart 2

+ – CH2 H :CH2

X:–

272

C.F. BERNASCONI

‘‘delocalized valence bond method’’ (DVB).115 The main conclusion, irrespective of which method was used, is that the triple ion structure f5 is dominant which is in full agreement with the computational results summarized in Table 14. Why are carbon-to-carbon proton transfers slow? Costentin/Sav eant analysis Applying DFT and QCISD methodology, Costentin and SavO˜ant126 have analyzed Equation (21) with Y=H, CH=CH2, NO2, and CH2=CH–NO2 using a different theoretical framework in order to answer the question: why are carbon-to-carbon proton transfers slow? A major difference in their approach is that their description of the reaction coordinate only requires consideration of the distance, Q, between the two fragments CH2Y and the intramolecular reorganization (i.e., delocalization)127 that occurs during the proton transfer, but not the distance, q, that defines the location of the proton as it moves along the reaction coordinate (Scheme 1). In other words, the dynamics of the reaction is entirely governed by the heavy atom reorganization although tunneling needs to be considered as well. The major conclusions emerging from this analysis are as follows. For the carbon-to-carbon proton transfer, the rate of the reaction depends on the intramolecular reorganization127 and the characteristics of the barrier through which the proton tunnels. The relative slowness of the proton transfer in a non-activated carbon system such as CH4/CH 3 is the result of a larger distance (Q) between the carbon centers compared to that between nitrogen or oxygen in the H2O/HO or NHþ 4 /NH3 systems. In the presence of an electronwithdrawing substituent such as a nitro group, there is an increase in the C– H bond polarity which leads to a decrease in barriers but this decrease is attenuated by the imbalance in the internal reorganization (charge localization-delocalization) which occurs during the proton transfer. The conclusions reached by Costentin and SavO˜ant are in fact quite consistent with our own. The main difference is that, according to these authors, ‘‘the notion of an imbalanced transition state should be placed within the context of charge localization-delocalization heavy-atom intramolecular reorganization rather than of synchronization (or lack thereof) between charge delocalization and proton transfer.’’

Q YCH2

H q

Scheme 1

CH2Y

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

273

Gas phase acidities and reaction barriers Table 16 provides a summary of the enthalpic gas phase acidities (DHo) and enthalpic reaction barriers (DH‡) for the reactions of Equation (21).117,118,120 The table includes data for the CH4/CH 3 reference system. Broadly speaking, the acidities mainly reflect the resonance stabilization of the anion, although the field effect of the Y-group adds significantly to the stabilization of the anion and, for Y = CN, it is the dominant factor. An attempt to quantify the relative contributions of the resonance and field effects as well as the potential role played by the polarizability of the Y-group was made by correlating DDHo = DHo(CH3Y) – DHo(CH4) with the gas phase substituent constants71 sF (field effect), sR (resonance effect), and sa (polarizability effect) according to the Taft128 Equation (28). DDH o ¼ roF sF þ roR sR þ roa sa

ð28Þ

The least squares correlation is shown in Fig. 3 for those systems for which sF, sR, and sa were available; the correlation was excellent (r2 = 0.992) and yielded roF = –43.0, roR = –192.5, and roa = –4.64 (Table 17). The ro values confirm the dominance of the resonance effect for most cases as well as the significant contribution of the field effect. They also suggest that the Table 16 Gas phase acidities of CH3Y and NCCH2Y (DHo) and reaction barriers (DH‡)a Y

H CH=CH2 C CH CH=NH (anti) CH=NH (syn) CN CH=O NO2 NO CH=S CH= OHþ NOþ 2H

CH3Y/CH2=Yb DHo

DH‡

DHo

DH‡

418.1 390.2 385.9 379.8 376.1 375.4 367.2 359.0 351.9 348.7 195.0 192.6

8.1 4.7 1.8 2.9 0.3 –8.5 –0.3 (1.5)d,e –6.2 (9.8)d,f –1.1 0.3 –5.1 –1.0

375.4 354.9

–8.5 –10.6

336.3 335.5 328.3 322.5 322.1

–14.3 –8.4 –16.3 –4.7 –7.0

In kcal mol1. From Reference 118. c From Reference 140. d Geometrically constrained transition state. e From Reference 113. f From Reference 117. a

b

NCCH2Y/ NCCH=Yc

274

C.F. BERNASCONI 430 420

CH4

410

ΔHo, kcal mol–1

400 390

CH3CH = CH2

380 CH3CN

370

CH3CH = O

360

CH3NO2 CH3NO

350 340 0

10

20 30 40 50 60 –43.0σF – 192.5σR – 4.64σα

70

80

Fig. 3 Plot of DHo according to Equation (28) for the acidities of CH3Y. Table 17 Analysis of acidities and barriers by means of Taft equations CH3Y/CH2=Ya

NCCH2Y/NCCH=Yb

DDHo roF roR roa

–43.0 –192.5 –4.64

–41.1 –135 0.54

DDH‡ r‡F r‡R r‡a

–22.6 9.81 7.59

–7.01 36.6 11.9

r

a b

From Reference 118. From Reference 140.

polarizability effect is almost negligible which is perhaps surprising in view of its potential importance for certain gas phase anions.129,130 The small role played by polarizability in the present systems suggests that its effect is greatly diminished when the ionic charge mainly resides on the Y-group rather than the neighboring CH2 group, as is the case for all present anions except for CH2CN. In fact, because the sa values are defined as negative numbers,128

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

275

the slightly negative roa value suggests a small anion destabilizing effect; however, in view of the smallness of the effect, not much significance should be attached to this finding. A possible interpretation of the negative roa value is that it could result from a slight stabilization of the neutral CH3Y. A clearer indication of the absolute and relative contributions of field, resonance, and polarizability effects to the acidity of the various compounds can be obtained by calculating the individual roF sF, roR sR, and roa sa terms for each acid rather than just focusing on the roF , roR , and roa values, respectively. These terms are summarized in Table 18; for the compounds with Y-groups with unknown substituent constants (Y = C CH, CH=NH, and CH=S), these terms were calculated based on approximate substituent constants estimated as described in reference 118. The information in Table 18 is quite revealing. For example, the resonance contribution amounts to about –63 kcal mol1 for CH3CH=S, –50 kcal mol1 for CH3NO, is in the range of –26 to –36 kcal mol1 for CH3CH=CH2, CH3C CH, CH3CH=NH, CH3CH=O, and CH3NO2, and a mere –19 kcal mol1 for CH3CN. On the other hand, the field effect contribution for CH3CN (–26 kcal mol1) is larger than for any of the other compounds except for CH3NO2 (–27 kcal mol1), and for CH3C CH (–6 kcal mol1) and CH3CH=CH2 (–2.6 kcal mol1) it is very small to almost negligible. The polarizability effect, if there is any significance to it at all, is seen to lower the acidities by a mere 2.1–3.4 kcal mol1 for CH3CH=CH2, CH3CN, and CH3CH=O, and even less (1.2 kcal mol1) for CH3NO2 and CH3NO. Turning to the barriers we note that they are defined as the difference in enthalpy between the transition state and the separated reactants. This is important because in gas-phase ion-molecule reactions the transition state is typically preceded by an ion–dipole complex131–133 formed between the reactants, and the term ‘‘barrier’’ is sometimes used for the enthalpy difference between the transition state and the ion–dipole complex. However, these ion– dipole complexes have little relevance to the main topic discussed in this chapter and hence the chosen definition of DH‡ is more appropriate. For reasons explained elsewhere,118 the barriers reported in Table 16 have not been corrected for the basis set superposition error (BSSE),134 although such corrected values are available.118 The barriers for all CH3Y/CH2=Y systems are lower than for the CH4/ CH 3 system. This means that the stabilization of the transition states by the Ygroup is greater than that of the respective anions. The situation is illustrated in Fig. 4 for the case of CH3NO2; it shows that the transition state for the CH3NO2 reaction is more stable than the transition state for the methane  reaction by 73.3 kcal mol1 while CH2=NO 2 is more stable than CH3 by only 1 59.1 kcal mol . The greater stabilization of the transition state compared to that of the anion may be attributed to the fact that, because the proton in flight is positively charged, each CH2Y fragment carries more than half a negative

276 Table 18 Dissection of the contribution of field, resonance, and polarizability effects to DHo and DH‡ in the CH3Y/CH2=Y systemsa CH3Y

sF

sR

sa

DDHo [Equation (28)] roF sF

roR sR

DHo

b

DDH‡ [Equation (29)] r‡F sF,

roa sa

r‡R sR

DH‡

b

r‡a sa

CH4 CH3CH=CH2 CH3CN CH3CH=O CH3NO2 CH3NO

Y-groups with 0 0 0.06 0.16 0.60 0.10 0.31 0.19 0.65 0.18 0.41 0.26

known sF, sR, and sa values 0 0 0 0 –0.50 –2.6 –30.8 2.3 –0.46 –25.8 –19.2 2.1 –0.46 –13.3 –36.6 2.1 –0.26 –27.9 –34.6 1.2 –0.25 –17.6 –50.0 1.2

418.1 390.2 375.4 367.2 359.0 351.9

(418.6) (387.5) (375.7) (370.8) (357.2) (352.1)

0 –1.4 –13.6 –7.0 –14.7 –9.3

0 1.6 1.0 1.9 1.8 2.6

0 –3.8 –3.5 –3.5 –2.0 –1.9

8.05 (8.08) 4.65 (4.50) –8.46 (–7.98) –0.31 (–0.55) –6.15 (–6.81) –1.06 (–0.53)

CH3C CH CH3CH=NH(syn) CH3CH=NH(anti) CH3CH=S

Y-groups with 0.14 0.15 0.27 0.17 0.20 0.17 0.22 0.33

estimated sF, sR, –0.40 –6.0 –0.45 –11.6 –0.40 –8.6 –0.75 –9.5

and sa valuesc –28.9 1.8 –32.7 2.1 –32.7 1.8 –63.5 3.4

385.9 376.1 379.8 347.7

(385.5) (376.4) (379.1) (349.0)

–3.2 –6.1 –4.5 –5.0

1.5 1.7 1.7 3.2

–3.0 –3.4 –3.0 –5.7

1.75 0.30 2.90 0.32

(3.36) (0.28) (2.28) (0.58)

From Reference 118, in kcal mol1. Number in parentheses from correlation according to Equation (28) (DHo) with DDHo = –43.0sF – 192.5sR – 4.60sa or Equation (29) (DH‡) with DDH‡ = –22.6sF – 9.81sR þ 7.59sa, respectively. c sF, sR, and sa estimated as described in Reference 118. a

C.F. BERNASCONI

b

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

(H3C

277

CH3)–

H

8.0 CH4 + CH3–

–59.1

CH2NO2 + CH2

–73.3

–

NO2

–6.2 (O2NCH2

H

CH2NO2)–

 Fig. 4 Stabilization of CH2=NO 2 relative to CH3 and stabilization of the transition state of the CH3NO2/CH2=NO reaction relative to the transition state of the CH4/ 2 CH 3 reaction.

charge (Table 14). This leads to a stronger substituent effect on the transition state than on the anion that results from the interaction of the Y-group with the negative charges. It also provides additional stabilization by electrostatic/ hydrogen bonding effects between the proton in flight and the negative CH2Y fragments.135 This latter effect is consistent with Gronert’s139 findings of an inverse correlation between transition state energy and charge on the transferred proton in identity proton transfers of nonmetal hydrides. The importance of electrostatic/hydrogen bonding effects can also be seen by comparing the barrier of the reactions going through the geometrically constrained transition state with that going through the optimized transition  state for the CH3NO2/CH2=NO 2 and CH3CH=O/CH2=CH–O systems. In the former system the barrier going through TS(constr) is 16.8 kcal mol1 higher than going through TS(optim), while for the CH3CH=O/CH2=CH– O system the difference is 10.5 kcal mol1 (Table 16). Using the VB approach mentioned earlier, Harris et al.119 calculated a 8.2 kcal mol1 higher energy for the delocalized transition state relative to that of the optimized structure in the

278

C.F. BERNASCONI

CH3CH=O/CH2=CH–O system. The higher energy of TS(constr), despite the larger resonance effect, probably results mainly from the fact that the product of the positive charge on the proton-in-flight and the negative charge on the CH2Y fragments is smaller for TS(constr) than for TS(optim) (Table 14) which greatly reduces the electrostatic/hydrogen bonding stabilization. Furthermore, to the extent that more resonance delocalization into the Ygroup occurs, the field effect of Y is reduced. The barriers for the CH3CH=OHþ /CH2=CHOH and CH3 NOþ 2 H/ CH2=NO2H systems fall within the same general range as for the other systems. This seems surprising since the much higher carbon acidities of CH3CH=OHþ and CH3 NOþ 2 H might have been expected to lead to much lower barriers. The most important reason for the higher than expected barriers is likely to be the absence of the stabilizing electrostatic and hydrogen bonding effects found in the CH3Y/CH2=Y systems that arise from the interaction of the positively charged proton-in-flight with the negative CH2 groups and/or the entire CH3Y fragments at the transition state. In the CH3 YHþ /CH2=YH systems, the electrostatic stabilization is not only lost but even replaced by a destabilization since the CH2YH fragments in the transition state are positively charged and this is expected to lead to a substantial increase in the barrier. Further insights were obtained by analyzing the relative contributions of field, resonance, and polarizability effects to the barriers in a similar way as for the acidities, i.e., by correlating DDH‡ = DH‡(CH3Y) – DH‡(CH4) with the respective Taft substituent constants according to Equation (29). The correlation is shown in Fig. 5; it yielded r‡F = –22.6, r‡R = 9.81 and r‡a = 7.59 with DDH ‡ ¼ r‡F sF þ r‡R sR þ r‡a sa

ð29Þ

r2 = 0.995 (Table 17). The r‡ values indicate that the field and polarizability effects lower the barriers while the resonance effect increases the barriers. The individual r‡F sF, r‡R sR, and r‡a sa terms which allow a detailed assessment of the relative contribution of the various effects to the barriers for each system are included in Table 18. The following conclusions emerge. 1. For all systems except CH3CH=CH2 and CH3CH=S, the field effect is dominant and lowers the barrier by substantial amounts (–7.0 to –14.7 kcal mol1). The barrier-lowering effect results from the fact that the transition state stabilization corresponds to roF sF þ r‡F sF, i.e., the field effect on the transition state is (roF þ r‡F )/roF = (–43.0 – 22.6)/ (–43.0) = 1.52-fold stronger than on the anion. 2. The polarizability effect contributes –1.9 to –5.7 kcal mol1 to the lowering of the barrier. For most cases this is a minor contribution compared to that of the field effect; for CH3C CH and CH3CH=S it is comparable to the

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

279

10 CH4

8 6

CH3CH = CH2

ΔH ‡ , kcal mol–1

4 2 0 CH3NO

CH3CH = O

–2 –4 –6

CH3NO2

–8

CH3CN

–10 –5

0

5 10 –22.6σF + 9.81σR + 7.59σα

15

20

Fig. 5 Plot of DH‡ according to Equation (29) for the CH3Y/CH2=Y systems.

field effect while for CH3CH=CH2 it is the dominant factor. The lowering of the barrier by the polarizability effect comes about because the stabilization of the transition state, which is given by roa sa þ r‡a sa, more than offsets the destabilization of the anion ( roa = –4.6, r‡a = 7.59). The fact that the transition states, which have less charge on the Y-group than the anions, are stabilized but that the anions are destabilized (or hardly affected) supports our conclusion mentioned above that the polarizability effect is greatly reduced when the Y-group carries a large negative charge. Because even for the transition states there is a significant charge on the Ygroup, the polarizability effect on their stability is still rather modest. This  contrasts with the reactions of the type ZCH3 þ ZCH 2 ! ZCH2 þ ZCH3 with Z = F, Cl, Br, OH, SH where the polarizability of Z has a strong barrier-reducing effect.130 3. The resonance effect increases the barriers by 1.0 to 3.2 kcal mol1. The reason for this small increase is that the resonance stabilization of the transition state which is given by roR sR þ r‡R sR is only ( roR þ r‡R )/ roR = (–192.5 þ 9.81)/(–192.5) = 0.95 as strong as that of the anion. Note that the rather modest barrier-enhancing effect of resonance is consistent with the rather small transition state imbalances (Table 15).

280

C.F. BERNASCONI

THE NCCH2Y/NCCH=Y SYSTEMS

The reactions of Equation (30) show many similarities with those of Equation (21) but there are also important Y CH2 CNþNCCH ¼ Y Ð Y ¼ CHCNþNCCH2 Y

ð30Þ

differences resulting from the strong electron-withdrawing effect of the cyano group. The specific systems studied include Y = CN, CH=CH2, CH=O, NO2, NO, and CH=S.140 The changes in the C–Y bond lengths that result from the ionization of NCCH2Y are summarized in Table 13. They are very similar to those for the ionization of CH3Y. On the other hand, the anionic charges on the Y-groups of NCCH=Y are significantly smaller than for CH2=Y (Table 13). This is because part of the charge is delocalized into the cyano group; this latter charge varies from –0.175 to –0.267 depending on Y. The transition states for the NCCH2Y/NCCH=Y systems are more imbalanced than those of the respective CH3Y/CH2=Y systems. This is seen both in the geometric parameters and the n values summarized in Table 15. In each case the C–Y contraction and planarization of the a-carbon at the transition state is less advanced than for the respective CH3Y/CH2=Y systems while n is larger, indicating a greater lag in charge delocalization in the NCCH2Y/ NCCH=Y systems. The larger imbalance has been attributed to the strong field effect of the cyano group which, because of its proximity to the a-carbon, strongly stabilizes the negative charge on that carbon. This allows for a greater accumulation of the negative charge on the a-carbon of the transition state in the NCCH2Y/NCCH=Y systems than in the CH3Y/CH2=Y systems. As was observed for the CH3Y/CH2=Y systems, the proton that is being transferred carries a significant positive charge. For the NCCH2Y/ NCCH=Y systems this charge is about 0.30  0.03, slightly larger than the 0.275  0.025 charge for the CH3Y/CH2=Y systems (Table 14). The slightly larger positive charge in the NCCH2Y/NCCH=Y systems may be related to the larger negative charges on the NCCHY fragments and provide greater electrostatic stabilization of the transition state, a point to be elaborated upon when discussing the barriers. The acidities of NCCH2Y and barriers of the NCCH2Y/NCCH=Y systems are summarized in Table 16. As expected, the acidities of NCCH2Y are substantially higher than those of CH3Y since the respective anions are strongly stabilized by the cyano group. The acidifying effect of the cyano group decreases as the p-acceptor strength of the Y-group increases. Analysis of the acidities according to Equation (31) affords roF = –41.1, roR = –135.0, and DDH o ¼ DH o ðNCCH2 YÞ  DH o ðNCCH3 Þ ¼ roF sF þ roR sR þ roa sa

ð31Þ

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

281

roa = 0.54 (Table 17). The roF and roa values are very similar to those observed for CH3Y but roR is substantially smaller compared to that for CH3Y. This means that the presence of the cyano group mainly reduces the resonance contribution of Y to the stabilization of the anion but not the contribution by the field and polarizability effects. The reduced roR value is consistent with the reduced charge on the Y-group (Table 13). Regarding the barriers, the cyano group leads to a substantial reduction in all cases, with the largest effect being on those CH3Y/CH2=Y systems with a relatively high barrier. This barrier reduction indicates that the electronwithdrawing effect of the cyano group stabilizes the transition state to a greater extent than the anion which is consistent with the larger transition state imbalance. This effect is reminiscent of the influence of electronwithdrawing substituents on DG‡o in solution phase reactions discussed in the section on ‘‘Polar effect of adjacent substituents.’’ Of particular interest is a comparison of the relative contributions of field, resonance, and polarizability effects to the barriers for the NCCH2Y/ NCCH=Y systems relative to those for the CH3Y/CH2=Y systems. These relative contributions were obtained from the correlation according to Equation (32) which yields r‡F = –7.01, r‡R = 36.6, and r‡a = 11.9 (Table 17). These

DDH ‡ ¼ DH ‡ ðNCCH2 YÞ  DH ‡ ðNCCH3 Þ ¼ r‡F sF þ r‡R sR þ r‡a sa

ð32Þ

numbers show the same qualitative pattern as for the CH3Y/CH2=Y systems, i.e., the field and polarizability effects lower the barriers while the resonance effect increases the barriers. But there are major quantitative differences. The field effect is much smaller than for the CH3Y/CH2=Y systems. This is because the cyano group takes over much of the role played by the field effect of the Y-groups. The barrier enhancement by the resonance effect is much greater than in the CH3Y/CH2=Y systems. As stated before, a positive r‡R value does not mean that the transition state is destabilized by the resonance effect; it only means that the resonance stabilization of the transition state is weaker than that of the anion. In quantitative terms, transition state resonance stabilization is given by roR sR þ r‡R sR which yields (roR sR þ r‡R sR)/roR sR = (roR þ r‡R )/ roR = (–135 þ 36.6)/(–135) = 0.73 as the fraction of transition state stabilization relative to resonance stabilization of the anion. This compares with (roR þ r‡R )/ roR = 0.95 for the CH3Y/CH2=Y systems. The smaller resonance stabilization of the transition state in the NCCH2Y/NCCH=Y systems is a direct consequence of the larger imbalance.

282

C.F. BERNASCONI

AROMATIC AND ANTI-AROMATIC SYSTEMS

In keeping with the advantages of examining identity reactions, a number of identity proton transfers involving aromatic systems were subjected to ab initio calculations. The first study involved the highly aromatic benzene and cyclopentadienyl systems, Equations (33a) and (34a).141

H

H

H

+

+

+

+

H

(33a)

+

+

+

+

(33b)

+

–

–

+

(34a) H

H

H

+

–

–

+

H

(34b)

Calculations at the MP2/6-311þG(d,p) level showed a significantly lower ‡ DH‡ for the more aromatic C6 Hþ 7 /C6H6 system (Equation (33a), DH = –7.6 1  kcal mol ) compared to the less aromatic C5H6/C5 H5 system (Equation (34a), DH‡ = 2.2 kcal mol1). A substantial lowering of the intrinsic barrier due to aromaticity was also deduced from a comparison between the DH‡ values for the aromatic systems and those for the corresponding noncyclic reference systems, i.e., 33a versus 33b, and 34a versus 34b. The numerical results of these calculations are summarized in Table 19. These results imply a disproportionately large aromaticity development at the transition state,

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

283

Table 19 Barriers (DH‡) of reactions 34a, 34b, 35a, 35b, 36a, and 36b and aromatic stabilization energies (ASE) in the gas phasea System

Equations

C6 Hþ 7 =C6 H6 +

C5 H6 =C5 H 5 –

C4 Hþ 5 =C4 H4 +

ASE (kcal mol1)

DH‡b (kcal mol1)

DDH‡c,d (kcal mol1)

33a 33b

–36.3

–7.6 3.5

–11.1

34a 34b

–29.4

2.2 9.8

–7.6

35a 35b

38.9

3.6 –2.4

6.0

a

At MP2/6-311þG**, Reference 141. Corrected for BSSE. c DDH‡ = DH‡(cyclic) – DH‡(noncyclic). b

i.e., the sum of the aromatic stabilization energies (ASEs) of the two halves of the transition state is greater than the ASEs of the respective aromatic reactant/product. This is illustrated by the schematic energy profiles shown in Fig. 6 for reactions 33a/33b (a) and 34a/34b (b), respectively. The arrows pointing down represent the aromatic stabilization energies of the reactants (ASER), products (ASEP), and the transition state (ASETS), respectively. The greater than 50% aromaticity in both halves of the transition state is reflected in the fact that |ASETS| > |ASER| = |ASEP|. The conclusions based on energy calculations are supported by the calculation of aromaticity indices such as HOMA142,143 and NICS(1)144,145 values as well as the pyramidal angle of the transition state. The pyramidal angle, a, is defined as illustrated for the benzenium ion (59) and the transition state (60) for reaction 33a (B = benzene); this angle is 0 in the aromatic species.

B H

H α α

α H

H 59

60

The various indices are summarized in Table 20; they show that the change in these indices in going from reactants to the transition state is significantly greater than 50%.

284

C.F. BERNASCONI

(A)

+

+

+

+

ASEP

ASER HH

HH

ASETS +

+

+

+

(B) –

+

–

+

HH

HH +

–

ASER

ASETS

ASEP

–

+

(C) HH +

ASER +

+

HH

STRAIN

+

+

+

ASEP ASETS

+

+

Fig. 6 Reaction energy profile for reactions 34a/34b (A), 35a/35b (B), and 36a/36b (C). (A) and (B): Aromatic stabilization of the transition state is greater than that of benzene or cyclopentadienyl anion, respectively. (C): Anti-aromatic destabilization (positive ASE) of the transition state is less than that of cyclobutadiene; the high barrier results from the additional contribution by angular and torsional strain at the transition state.

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

285

Table 20 Aromaticity indices for the reactions of Equations (34a), (35a), and (36a)a HOMA C6 Hþ 7 TS C6H6 % Progress at TSb C5H6 TS C5 H 5 % Progress at TSb C4 Hþ 5 TS C4H4 % Progress at TSb a b

NICS(1)

Equation (34a) 0.415 0.874 0.963 83.8 Equation (35a) –0.791 0.560 0.739 88.3 Equation (36a) –0.99 –0.156 –3.55 22.3

–6.05 –9.26 –10.20 77.3 –0.517 –8.33 –9.36 75.4 –13.69 –12.64 18.11 3.30

a 50.1 11.5 0.0 71.0 53.5 21.9 0.0 59.0 60.5 52.2 0.0 13.7

At MP2/6-311þG**, Reference 141. [Index(TS) – Index(Reactant)]/[Index(Product) – Index(Reactant)]  100.

A reaction of particular interest is that of Equation (35a) because it involves an anti-aromatic system. The barriers for Equation (35a) and its noncyclic reference system [Equation (35b)] are included in H

H

+

+

H +

+

H +

(35a)

+

+

+

(35b)

Table 19 while the corresponding aromaticity indices are included in Table 20. The HOMA, NICS(1), and a values all indicate a very small degree of antiaromaticity development at the transition state.146 Since the transition states for Equations (33a) and (34a) are able to benefit from the stabilization conferred by the strong development of aromaticity it seems reasonable that in Equation (35a) the transition state should be able to avoid much of the destabilization that arises from anti-aromaticity, i.e., keeping the development of anti-aromaticity lagging behind proton transfer. Hence, according to the PNS, the barrier should be lowered by this effect. However, DH‡ for Equation

286

C.F. BERNASCONI

(35a) was calculated to be higher than for Equation (35b) (Table 19) which seems inconsistent with the above conclusions unless the barrier-lowering PNS effect is masked and overshadowed by other factors. As discussed in more detail elsewhere,141 angle and torsional strains at the transition state are in fact believed to be responsible for the higher than expected barrier, i.e., in the absence of these strains the barrier would indeed be lower than for Equation (35b). This is illustrated in Fig. 6, part (C). Another study involved the identity reactions shown in Equation (36).147 Calculations at the O–

O

O–

+

O

+

(36) X

X

X

X

61H-O (X = O)

61–-O (X = O)

61–-O (X = O)

61H-O (X = O)

61H-S (X = S)

61–-S

61–-S

61H-S (X = S)

(X = S)

(X = S)

MP2/6-31þG** level yielded a DH‡ for X = S that is lower than for X = O. It was also found that DH‡ for the reactions of Equation (36) was lower than for the corresponding noncyclic reference systems of Equation (37). The results are summarized in Table 21. Furthermore, the intrinsic barrier for O–

O CH2

CH

C

CH2XCH3 + CH2

CH

C

62H-O (X = O)

62–-O (X = O)

62H-S (X = S)

62–-S (X = S)

CHXCH3

(37) O– CH2 62–-O

CH

(X = O) 62–-S (X = S)

C

O CHXCH3 + CH2

CH

C

CH2XCH3

62H-O (X = O) 62H-S (X = S)

the deprotonation of 61H-S by CN is lower than for the deprotonation of 61H-O by the same base. All these results indicate that aromaticity lowers the intrinsic barrier and increasingly so with increasing aromaticity. Further evidence showing disproportionately high transition state aromaticity comes form NICS values,144,145 Bird indices,148,149 and HOMA142,143

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

287

Table 21 Barriers of reactions 37 and 38 in the gas phasea System

DH‡ (kcal mol1)

DDH‡b (kcal mol1)

3.6 5.4 2.3 4.3

–1.8

61H-O/61-O 62H-O/62-O 61H-S/61-S 62H-S/62-S a b

–2.0

At MP2/6-31þG**, Reference 147. DDH‡ = DH‡(cyclic) – DH‡(noncyclic).

Table 22 Aromaticity indices for the identity proton transfers of Equation (36)a Species

HOMA

Bird index

NICS(–1)

61H-O TS 61-O % Progress at TSb 61H-S TS 61-S % Progress at TSb

–0.686 0.217 0.544 73.4 –0.854 0.231 0.566 76.4

22.68 36.65 40.28 80.0 31.31 53.19 64.20 66.5

–2.47 –5.46 –6.74 69.9 –2.45 –5.37 –7.33 59.8

a b

At MP2/6-31þG**, Reference 147. [Index(TS) – Index(Reactant)]/[Index(Product) – Index(Reactant)]  100.

values as indicators of aromaticity. These aromaticity indices are summarized in Table 22 for the identity reactions [Equation (36)]. As was the case for reactions 33a and 34a, the progress in the development of aromaticity at the transition state is greater than 50%. NICS, HOMA, and Bird indices were also calculated for the transition states of the reactions of 61H-O and 61H-S with a series of carbanions. The results are reported in Table 23. The trends in these parameters show a clear increase as the transition state becomes more product-like with increasing endothermicity, indicating an increase in transition state aromaticity. Even more revealing is the % progress at the transition state which indicates that this progress is >50% not only for the endothermic reactions (product-like transition states) but even for most of the exothermic reactions (reactant-like transition states) except those with strongly negative DHo values. Additional confirmation of early development of aromaticity as the reaction progresses comes from plots of NICS values and Bird indices versus the reaction coordinate for the reaction of 61H-S with CH2 NO (Figs. 7 and 8), and of 61H-O with CH2 NO2 (figures not shown). These reactions were chosen

288

Table 23 Transition state aromaticity indices for the reactions of 61H-O and 61H-S with carbanions in the gas phasea % progress at TS 1

o

DH (kcal mol )

HOMA

Bird index

NICS(–1)

HOMA

Bird index

61H-Ob  CH2CN  CH2CO2H  CH2COCH3  CH2CHO  CH2NO2 CH3 CHNO2  CH(CN)2

–19.8 –14.3 –13.0 –10.4 –1.2 –0.9 19.8

–0.097 0.083 0.105 0.132 0.115 0.126 0.245

28.99 32.58 32.81 34.04 33.34 34.04 33.89

–3.98 –4.70 –4.64 –4.99 –4.92 –4.76 –5.04

47.9 62.6 64.4 66.6 65.2 66.1 75.8

–0.097 0.083 57.6 64.6 60.6 64.6 63.7

33.3 52.1 50.7 59.0 57.3 53.5 60.2

61H-Sc  CH2CN  CH2CHO  CH2NO2  CH2NO  CH2CHS  CH(CN)2  CH(NO2)2

–26.5 –17.1 –7.9 –0.3 1.3 13.1 24.7

–0.188 0.062 0.082 0.248 0.196 0.l36 0.236

42.66 48.19 48.91 54.17 53.79 51.31 53.08

–3.71 –4.69 –4.92 –5.43 –5.55 –5.03 –5.42

46.3 64.5 65.9 77.6 71.2 73.9 76.7

33.9 51.3 53.5 69.5 68.3 60.8 66.2

25.9 46.0 50.7 61.0 63.6 52.9 60.8

RCHY

At MP2/6-31þG**, Reference 147. 61H-O/61-O: HOMA – 0.686/0.544, Bird index 22.68/40.28, NICS(–1) – 2.47/–6.74. c 61H-S/61-S: HOMA – 0.854/0.566, Bird index 31.31/64.20, NICS(–1) – 2.45/–7.33. b

C.F. BERNASCONI

a

NICS(–1)

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

289

0 NICS(0)

–2

NICS(–1)

NICS value

–4

–6

–8

–10

–12 –2.0

–1.5 –1.0 –0.5

0.0

0.5

1.0

1.5

2.0

2.5

Reaction coordinate (amu½ Bohr)

Fig. 7 Plots of NICS(0) and NICS(–1) versus IRC for the reaction of 61H-S with CH2 NO.

70 65

Bird Index

60 55 40 35 30 25 20 –2.0

–1.5

–1.0

–0.5

0.0

0.5

1.0

1.5

2.0

2.5

Reaction coordinate (amu½ Bohr)

Fig. 8 Plot of the Bird index versus IRC for the reaction of 61H-S with –CH2NO.

290

C.F. BERNASCONI

because they are nearly thermoneutral (see Table 23) and have fairly symmetrical transition states as indicated by the C- --H- --B bond lengths.147 The plots show a steep rise in aromaticity as a function of the reaction coordinate as the transition state is reached and a pronounced leveling off toward the value of the anionic product once the transition state has been traversed. As indicated in Table 23, the % progress in the development of product aromaticity at the transition state of the reaction of 61H-O with CH2 NO2 is 57.3 for NICS(–1) and 60.6 for the Bird index, while for the reaction of 61H-S with CH2 NO these percentages are 61.0 and 69.5, respectively. Decoupling of aromaticity development from charge delocalization In solution phase reactions such as Equation (1) as well as in the gas phase reactions of Equation (21) charge delocalization always lags behind proton transfer at the transition state. For the solution phase reactions this feature not only manifests itself in enhanced intrinsic barriers but also in the Br½nsted coefficients. For the gas phase reactions this lag can be deduced from calculated NPA charges. An interesting question is whether in systems such as Equation (33a), (34a), or (36) the early development of aromaticity would induce charge delocalization to do the same rather than to follow the typical pattern of delayed charge delocalization found in non-aromatic systems. NPA charges for some representative systems are shown in Chart 3. They indicate that negative charge is being created at the reaction site of the transition state  which then either disappears (C6 Hþ 7 /C6H6 system) or decreases (C5H6/C5 H5 system) due to delocalization in the product. This implies that, in terms of

H

H

0.095

H

0.905

H

0.001

Chart 3

H –0.001

0.428 H –0.154

0.440

0.000

0.000 0.376 H

–0.200

H –0.275

–0.413

–0.800

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

291

charge delocalization, the transition state is imbalanced in the same way as in non-aromatic systems, i.e., development of aromaticity and charge delocalization are decoupled. Comparisons with aromatic transition states in other reactions Aromaticity in transition states is a well-known phenomenon, especially in pericyclic reactions, as recognized more than half a century ago.150–152 A prototypical example is the Diels–Alder reaction of ethylene þ 1,3-butadiene ! cyclohexene; a computational study by the Schleyer group153 has shown that the absolute value of the diamagnetic susceptibility, which is a measure of aromaticity, goes through a maximum at the transition state. Similar situations have been reported for other Diels–Alder reactions,154 for 1,3-dipolar cycloadditions,155 and enediyne cyclizations,156 where the aromaticity of reactants, products, and transition states was evaluated using NICS values. A recent report regarding transition state aromaticity in double group transfer reactions such as the concerted transfer of two hydrogen atoms from ethane to ethylene is also worth mentioning.157 For many additional examples and references, the review by Chen et al.145 should be consulted. It is important to note, though, that for these reactions the situation is quite different from that in Equations (33a), (34a), and (36). In pericyclic reactions aromaticity is mainly a special characteristic of the transition state whereas the reactants and products are not aromatic or less so than the transition state. This is quite different from the proton-transfer reactions discussed in this chapter where the aromaticity of the transition state is directly related to that of the reactants/products. An analogy with steric effects on reaction barriers may illustrate the point. In a reaction of the type of Equation (38), steric effects at the transition state will definitely increase the intrinsic A þ BÐC þ D

ð38Þ

barrier if the reactants are bulky. However, because there are no steric effects on the reactants or products, the concept of early or late development does not apply here, and the same is true for the aromaticity of the Diels–Alder transition state. In contrast, in a reaction of the type of Equation (39) there is steric crowding both in the product and the transition state. In this case, the intrinsic A þ BÐAB

ð39Þ

barrier will be enhanced if steric crowding has made disproportionately large progress relative to bond formation at the transition state, as is the case for

292

C.F. BERNASCONI

nucleophilic addition to alkenes discussed in a later section; on the other hand, DG‡o will be reduced if development of the steric effect is disproportionately small. This, then, is akin to early or late development of transition state aromaticity or anti-aromaticity in our reactions.

Aromaticity versus resonance Why does aromaticity and resonance affect intrinsic barriers differently? The lowering of the barrier by providing the transition state with excess aromatic stabilization appears to be in keeping with Nature’s principle of always choosing the lowest energy path. The fact that the transition states are able to be so highly aromatic suggests that only relatively minor progress in the creation of appropriate orbitals or the establishment of their optimal alignment and distances from each other may be required for aromatic stabilization to become effective. There are several precedents that support this notion. For example, the NICS value of KekulO˜ benzene (rCC fixed at 1.350 e´ and 1.449 e´) is only 0.8 ppm less than the NICS value for benzene itself or, with rCC = 1.33 e´ (ethylene-like) and 1.54 (ethane-like), the NICS value is only 2.6 ppm less than that for benzene.145,158 Or the NICS value for 63 (–8.1 ppm)159 is quite close to that of benzene (–9.7 ppm)145,158 even though there is strong bending of the benzene ring. Other relevant observations have been discussed elsewhere.141

CN

NC

CN

NC

63

In contrast, in reactions that lead to resonance stabilized/delocalized products such as Equation (1) or (21), the transition state is not able to maximize the potentially stabilizing effect of extensive charge delocalization. As discussed in the section ‘‘Why does delocalization lag behind proton transfer,’’ this is because delocalization can only occur if there is significant C–Y p-bond formation. Hence, the fraction of charge on Y at the transition state depends on the fraction of p-bond formation which in turn depends on the fraction of charge transferred from the base to the carbon acid. This imposes an insurmountable constraint on the transition state because the charge on Y can never be large since it is a fraction of a fraction [Equations (12) and (25)].

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

4

293

Other reactions

NUCLEOPHILIC ADDITIONS TO ALKENES

Reactions of the type of Equation (40) with nucleophiles such as HO, water, aryloxide ions, thiolate

Y Nuν +

C

C

Y–

k1 k–1

C

C

(40)

Nuν + 1

ions, and amines follow similar patterns as proton transfers of the type of Equation (1). This is not surprising since delocalization of the negative charge into the Y-group plays a similar role as in Equation (1). Most of the systematic kinetic studies of these reactions were published before 1992 and thus have been reviewed in detail in our 1992 chapter.4 Hence, only a brief summary, based on Reference 4, of the major features is given here. The main conclusions can be summarized as follows. Correlation with proton transfers There is a strong correlation between the intrinsic rate constants of reactions 40 with those of reactions 1. For example, a plot of log ko for the reactions of piperidine and morpholine with PhCH=C(CN)2, PhCH=C(COO)2C(CH3)2, PhCH=C(CN)C6H4-4-NO2, PhCH=C(CN)C6H3-2,4-(NO2)2, PhCH=CHNO2, PhCH=C(C4Cl4),160 and PhCH=C(Ph)NO2 versus the log ko for the deprotonation of CH2(CN)2, CH2(COO)2C(CH3)2, 4-NO2-C6H4CH2CN, 2,4-(NO2)2C6H3CH2CN, CH3NO2, C5H2Cl4,160 and PhCH2NO2, respectively, gives a good linear correlation. This indicates that the resonance effect of the p-acceptors is qualitatively similar in both reactions. However, there is an attenuation of this effect in Equation (40) as indicated by the slope of 0.46. This attenuation is also reflected in smaller transition state imbalances, as measured by annuc  bnnuc ; annuc was obtained from plots of log k1 versus log K1 by varying the aryl substituents and corresponds to aCH in proton transfers, while bnnuc was obtained from plots of log k1 versus log K1 by varying the nucleophile and corresponds to bB in proton transfers. A major reason for the reduced imbalances is the fact that the b-carbon in the alkene is already sp2-hybridized which facilitates p-overlap with the Y-group at the transition state. This is symbolized in 64 by showing a small degree of charge delocalization into the Y-group as indicated by the small ‘‘d–.’’

294

C.F. BERNASCONI

C

Yδ–

δ– C

Nuν + δ 64

There are other factors that contribute to the reduction in the imbalance. This can be seen by comparing intrinsic rate constants of reactions that create the same carbanions as in proton transfer reactions, e.g., comparing reactions 41 and 42. These reactions do involve sp3 ! sp2 rehydrization just as in proton transfers and hence, if hydrization were the only important factor, PhCH O

CH(CN)2

PhCH

– O + CH(CN)2

CH2NO2

PhCH

O + CH2

(41)

–

PhCH

– NO2

(42)

O–

the difference in the log ko values between Equations (41) and (42) should be comparable to that between the log ko values for the deprotonation of malononitrile and nitromethane, respectively. The difference, Dlog ko  3.9, between reactions 41 and 42 is indeed larger than for the corresponding nucleophilic addition reactions to PhCH=C(CN)2 and PhCH=CHNO2, respectively (Dlog ko  2.6), but still smaller than for the corresponding proton transfers (Dlog ko = 6.8). A smaller Dlog ko (5.1) than for proton transfers was also found by the Crampton group161,162 for reactions 43 and 44. H

CH(CN)2 NO2

O2N

NO2

O2N

– + CH(CN2)

–

NO2

NO2

H

CH2NO2 NO2

O2N

NO2

O2N

+

–

NO2

(43)

NO2

CH2

– NO2

(44)

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

295

One potential second reason for the attenuation of the imbalance is steric hindrance to perfect co-planarity of Y in the anionic adducts of Equation (40). An additional factor that may enhance the differences between the imbalances in proton transfers and those in the other carbanion forming reactions is hydrogen bonding in the transition state of proton transfer. This hydrogen bonding stabilizes the transition state by keeping more of the negative charge on the carbon. The gas phase ab initio calculations discussed earlier support this notion.

Effect of intramolecular hydrogen bonding In reactions with amine nucleophiles the reaction leads to a zwitterionic adduct that is in rapid acid–base equilibrium with its anionic form, Equation (45). In some cases, the zwitterion is strongly

Y C

C

+ RR′NH

Y–

k1 C C + RR′NH

k–1

Y–

‡

Ka

H+

C

C

(45)

RR′N

stabilized by an intramolecular hydrogen bond as in the example of the reaction of benzylideneacetylacetone (65) with secondary amines. At the transition state this hydrogen bond

CH3 Ph H

C C

C

– C

RR′N+

O CH3

O H 65

is only weakly developed because the partial charges on the nitrogen and oxygen atoms are small and the distance between the donor and acceptor atoms is relatively large. Hence the stabilizing effect at the transition state is disproportionately small which leads to a reduction in the intrinsic rate constants of such reactions.

296

C.F. BERNASCONI

Steric effects Steric effects reduce rate and equilibrium constants of nucleophilic additions but the question how the intrinsic barrier is affected does not always have a clear answer. Comparisons of intrinsic rate constants for the addition of secondary alicyclic amines versus primary aliphatic amines suggest that ko is reduced by the F-strain. This implies that the development of the F-strain at the transition state is quite far advanced relative to bond formation. The effects of other types of steric hindrance on ko such as prevention of coplanarity of Y in the adduct or even prevention of p-overlap between Y and the C=C double bond in the alkene have not been thoroughly examined and hence are less well understood. Effect of polar substituents The polar effect of remote substituents on intrinsic rate constants is qualitatively similar to that in proton transfer, irrespective of whether the aryl group is attached to the a-carbon (e.g., 66) or the b-carbon (67); in both cases the partial negative charge in the transition state is closer to the Z

C

δ– Y C

δ– Y C

C

Nuν + δ

Z

Nuν + δ

66

67

substituent Z than in the adduct where the charge is on Y and hence an electron-withdrawing substituent will increase the intrinsic rate constant. This is also reflected in the above-mentioned fact that annuc > bnnuc . p-Donors in the para position of a-aryl groups can lead either to a reduction or an enhancement of the intrinsic rate constant. Examples where a reduction has been observed is the reaction of amines with benzylidene malononitriles, Equation (46), or with benzylidene Meldrum’s Z

Z CN C H

+ R2NH

C CN

CN

k1 k–1

H

C + R2NH

C CN

(46)

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

297

acid. This reduction may be understood in terms of the resonance stabilization of the alkene (68 ! 68). Just as is true for resonance effects in general, the loss of this reactant resonance Z

Z+ CN

CN C

C

C CN

H

H

C

– CN

68±

68

stabilization is expected to be ahead of C–N bond formation at the transition state which should decrease ko. In the reaction of piperidine with b-nitrostyrenes, Equation (47), ko for the reactions of the Z

Z H C H

C

H + R2NH

NO2

H

C + R2NH

(47)

C – NO2

p-OMe derivatives is enhanced, suggesting the above early loss of resonance is overshadowed by another effect. This effect can be understood as arising from a preorganization of the =CHNO2 group in the reactant toward its structure in the adduct (–CH=NO 2 ), thus reducing the transition state imbalance and avoiding some of the detrimental effect of the lag in the charge delocalization. This explanation is similar to that given for the p-donor effect in the deprotonation of Fischer carbenes and of 43-Z. Apparently the two types of p-donor effects operate simultaneously, with the former dominating in reaction 46 and the latter dominating in reaction 47. Polarizable nucleophiles Even though the high carbon basicity of thiolate ion nucleophiles is a major reason why their nucleophilic reactivity is much higher than that of oxyanions or amines of comparable pKa, there is an added effect that comes from a reduced intrinsic barrier. For example, intrinsic rate constants for thiolate ion addition to a-nitrostilbene or b-nitrostyrene are up to 100-fold higher than for amine addition. This has been explained in terms of the soft–soft interaction

298

C.F. BERNASCONI

responsible for the high thermodynamic stability of thiolate ion adducts developing ahead of C–S bond formation.

NUCLEOPHILIC VINYLIC SUBSTITUTION (SNV) REACTIONS

The most common mechanism of nucleophilic vinylic substitution163 is the two-step process of Equation (48) shown for the reaction of an anionic nucleophile with a vinylic substrate activated by one

Y

R Nu–

C

+

C Y′

LG

R k1

LG

C

k–1

Y

k2

C

Y

R C

C

Nu

Y′

+ LG– Y′

(48)

Nu 69

or two electron-withdrawing substituents Y and/or Y0 and a leaving group LG.163–170 The first step is essentially the same as that in nucleophilic additions to alkenes, Equation (40), except that steric and electronic effects of the leaving group affect the reactivity not only of the k2 step but also of the k1 and k–1 steps in important ways. Early mechanistic work on these reactions dealt exclusively with systems where 69 is an undetectable steady-state intermediate,163–167 making it impossible to determine intrinsic rate constants. However, more recent studies focusing on systems with strong nucleophiles such as thiolate or alkoxide ion, poor leaving groups such as alkoxide or thiolate ions, and strongly activated vinylic substrates allowed direct observation of 69 and determination of the individual rate constants k1, k–1, and k2.168,170 The reactions of 70-LG-74-LG with O

O NO2

Ph C

Ph

C

LG

O

CH3

Ph

CH3

LG

C Ph

C

LG

O O

70-H (LG = H) 70-OMe (LG = OMe) 70-Pr (LG = SPr-n) Ph C LG

Ph

NO2

LG

C

CN C

73-SMe (LG = SMe)

O

71-H (LG = H) 71-OMe (LG = OMe) 71-SMe (LG = SMe)

COOMe

C

C CN

74-H (LG = H) 74-OMe (LG = OMe)

72-H (LG = H) 72-SMe (LG = SMe)

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

299

thiolate ions171–176,178 have provided the most insights into how structural factors affect intrinsic rate constants in such systems. Table 24 summarizes  RS RS kRS 1 , K1 , and k2 values for the reactions with HOCH0 2CH2S as a repreCH2 YY sentative thiolate ion. Included in the table are pKa values of CH2YY0 ,  RS log ko for the intrinsic rate constants for RS addition determined from plots PT of log kRS versus log KRS values for the deprotonation of 1 1 , and log ko 0 CH2YY by secondary alicyclic amines. CH2 YY0 Figure 9 shows an excellent correlation of log KRS for 1 with pKa 0 LG = H( ) (slope = 1.11), indicating charge stabilization by YY in the adduct is similar to that for CHYY0 . For LG = OMe () and SMe (D) the correlation is poor due to steric crowding in the adduct which is strongest for YY0 = MA,177 intermediate for YY0 = ID,177 (NO2,CO2Me) and (Ph,NO2), and smallest for YY0 = (CN)2. The leaving group steric effects follow the expected order SPr-n > SMe > OMe >> H. Figure 10 shows that the correlations between log kRS and log KRS are 1 1 RS poor, implying that ko differs substantially from substrate to substrate and not only depends on YY0 but on the leaving group as well. This is best demonstrated in Fig. 11 which shows that, for a given leaving group, there is a linear correlation between log kRS and log kPT just as had been o o observed for the correlation between log ko for the reaction of amines with alkenes [Equation (40)] and log ko for the corresponding proton transfer mentioned earlier. Also as observed for reaction 40, the slopes are less than unity (0.32 for LG = H, 0.40 for LG = OMe, and 0.56 for LG = SMe) due to the sp2 hybridization of the b-carbon which facilitates overlap with the YY0 groups at the transition state and reduces the imbalance. However, as the differences in the slopes imply, the degree by which the imbalance is reduced depends on the leaving group and is largest for LG = H and smallest for LG = SMe. This conclusion is corroborated by the Br½nsted-type coefficients annuc and bnnuc for the reactions of thiolate ions with the phenyl-substituted Meldrum’s acid derivatives of 71-H and 71-SMe: for 71-SMe, annuc – bnnuc = 0.34, for 71-H annuc – bnnuc = 0.13, implying a smaller imbalance for 71-H. We further note that the kRS o values for the reactions with LG = OMe and SMe are much lower than with LG = H, especially for LG = SMe. There are two main factors that contribute to this result. One is the p-donor effect of the OMe and SMe groups (75 $ 75) which reduces kRS o

Y

R MeX

+ MeX

Y′ 75

Y

R

C

C

X = O or S

C

C Y′ 75±

300

Table 24 Rate and equilibrium constants for SNV reactions with HOCH2CH2Sin 50% DMSO–50% water at 20C 2 YY pKCH a

Substrate Ph

0

1 1 kRS s ) 1 (M

1 KRS 1 (M )

1 kRS 2 (s )

log kRS o

log kPT o

References

CN C

C

H

(74-H)

10.21

4.40  106

5.18  104

5.7

7.0

171

(70-H)

7.90

5.18  104

8.16  106

3.4

–0.25

172

(72-H)

6.35

4.47  106

1.16  109

4.8

3.13

171

(71-H)

4.70

1.44  107

5.38  1010

5.2

3.90

173

10.21

2.80  105

1.62  102

5.1

70

171

CN

Ph

Ph C

C

H

NO2 O

Ph C

C

H O O O CH3

Ph C

C

O CH3

O Ph

CN C

MeO

(74-OMe)

C CN

0.133

C.F. BERNASCONI

H

C

C

MeO

NO2

(70-OMe)

7.90

3.89  102

7.59  103

9.60  106

2.2

–0.25

174

(71-OMe)

4.70

4.40  104

2.57  104

2.16  104

3.7

3.13

175

(70-SPr)

7.90

4.70

10.4

4.50  102

0.29

2.44

175

(72-SMe)

6.35

5.62  102

2.25  102

0.245

2.5

3.90

171

(73-SMe)

5.95

2.48  102

5  104

5.80  105

£1.1

171

(71-SMe)

4.70

9.22  102

3.32  102

0.115

2.5

176

Ph O O CH3

Ph C

C MeO

O CH3

O Ph

NO2 C

C

n-PrS

Ph O

Ph C

C

MeS O Ph

Ph C

C

MeS

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

Ph

CO2Me O O CH3

Ph C

C

MeS O

O CH3

301

302

C.F. BERNASCONI 12

71-H 10

72-H

log K1RS

8

70-H 6

4

74-H

73-H 71-OME 70-OME

2

72-SMe 71-SMe

74-OME

70-SR 0 –12

–10

–8

–6

–4

–2

0

CH YY′ –pKa 2 0

2 YY Fig. 9 Plots of log KRS (RS = HOCH2CH2S) versus pKCH . a 1 LG = OMe; D, LG = SMe.

,

LG = H;

,

by the PNS effect of the early loss of the resonance stabilization of the substrate; this is similar to the effect of p-donor substituents in the phenyl group of alkenes as, e.g., in 68 $ 68. In view of the stronger p-donor effect of the OMe group,71 this factor should affect the reactions with LG = OMe more strongly than those with LG = SMe. However, since kRS o for the MeS derivatives is lower than for MeO derivatives, there must be one or more additional factors that reduce kRS o for the MeS derivatives relative to that for the MeO derivatives. One such factor appears to be steric hindrance (Fstrain) at the transition state which is quite advanced relative to the C–S bond formation and hence should result in a greater reduction of ko for the MeS derivative due to the larger size of the sulfur atom. This conclusion is in agreement with one reached for the reaction of amines with alkenes discussed earlier. Another factor may be the stronger electron-withdrawing inductive effect of the MeO group which imparts greater stabilization to the imbalanced transition state than the MeS group and hence increases the ko(OMe)/ko(SMe) ratio.

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

303

8

71-H 72-H

74-H 6

log k 1RS

74-OMe 70-H

71-OMe 4

71-SMe 70-OMe

→

72-SMe

73-SMe

2

70-SPr 0

0

2

4

6

8

10

12

RS

log K1

  RS Fig. 10 Plots of log KRS 1 versus log K1 (RS = HOCH2CH2S ) generated by varying YY0 . , LG = H; , LG = OMe; D, LG = SMe.

NUCLEOPHILIC SUBSTITUTION OF FISCHER CARBENE COMPLEXES

Intrinsic rate constants The reactions of Fischer carbene complexes with an anionic nucleophile may be represented by Equation (49).179–181 Typical carbene complexes that have been the subject of kinetic studies are 76-M and XR Nu– + (CO)5Cr

C R′

77-M

182–185

XR (CO)5Cr

k –1

k2

R′

(CO)5Cr

+ RX–

C

(49)

R′

as well as others mentioned below.

C

OCH2CH2O–

SMe (CO)5M

Ph

76-Cr (M = Cr) 76-W (M = W)

C Nu

OMe (CO)5M

Nu

–

k1

(CO)5Cr

C Ph

77-Cr (M = Cr) 77-W (M = W)

C

SCH2CH2O– (CO)5Cr

C

Ph 78-Cr

Ph 79-Cr

304

C.F. BERNASCONI 6

74-H

71-H 5

72-H 74-OMe

log k ORS

4

71-OMe 70-H

3

72-SMe

2

71-SMe

70-OMe

73-SMe

1

70-SPr 0 –1

0

1

2

3 log

4

5

6

7

8

PT kO

  PT Fig. 11 Plots log kRS o (RS = HOCH2CH2S ) versus log ko . , LG = H; , LG = OMe; D, LG = SMe.

These reactions show many similarities with the SNV reactions of Equation (48) but there are differences as well. Table 25 summarizes approximate log ko values for the addition of various nucleophiles to 76-M and 77-M. The table includes results for the intramolecular reactions of 78-Cr186 and 79-Cr186 that lead to the cyclic intermediates 80-Cr and 81-Cr, respectively. For

– (CO)5Cr

O C Ph 80-Cr

– O

(CO)5Cr

S C

O

Ph 81-Cr

the purpose of comparison, log ko values for the reactions of the vinylic substrate with the highest intrinsic rate constant (74-OMe) and the ones with

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

305

Table 25 Approximate intrinsic rate constants for the reactions of Fischer carbene complexes with nucleophilesa log ko 

Nu = MeO

b



Nu = ROc

Nu = RSc

OMe (CO)5Cr

C

(76-Cr)

0.96d

0.74e

2.11f

(76-W)

1.25d

0.96e

2.56f

Ph OMe (CO)5W

C Ph OCH2CH2O–

(CO)5Cr

(78-Cr)

C

1.19g

Ph SMe (CO)5Cr

C

(77-Cr)

–0.3h

(77-W)

0.0h

Ph SMe (CO)5W

C Ph SCH2CH2O–

(CO)5Cr

(79-Cr)

C

–1.53g

Ph OMe

NC C

C

NC C

C

(70-OMe)

2.2i

(70-SPr)

0.29i

Ph

Ph

SPr-n

O 2N C

a

5.1i

OMe

O 2N

Ph

(74-OMe) Ph

C Ph

In most cases log ko was determined using the simplest version of the Marcus equation, log ko = log k1 – 0.5 log K1. b In methanol at 25C. c In 50% MeCN–50% water (v/v) at 25C. d Reference 182. e Reference 183. f Reference 184. g Reference 186. h Reference 185. i From Table 24.

306

C.F. BERNASCONI

the lowest ko values (70-OMe and 70-SMe) are included in the table. The following points are noteworthy. 1. The intrinsic rate constants for thiolate ion addition to the Fischer carbenes are close to those for thiolate addition to the respective a-nitrostilbene derivatives 70-OMe and 70-SMe but much lower than for thiolate ion addition to methoxybenzylidinemalononitrile (74-OMe); for example, log ko = 2.1 for 76-Cr versus log ko = 2.2 for 70-OMe versus log ko = 5.1 for 74-OMe, or, log ko = –0.3 for 77-Cr versus log ko = 0.29 for 70-SPr. This is consistent with the extensive charge delocalization into the (CO)5M moiety that is responsible for the relatively high stability of the addition complexes176,177,183 and the lag of this delocalization behind bond formation at the transition state, see, e.g., 82; note that to show the imbalance the partial negative charge is placed on the metal atom rather than on the entire (CO)5M group. OMe δ– (CO)5M

C

Ph

δ– Nu 82

2. The intrinsic rate constants are much lower for nucleophilic attack on the thia carbene complexes than on the oxa carbene complexes. This is true irrespective of the nucleophile. For example, log ko = –0.3 for RS attack on 77-Cr versus log ko = 2.1 for RS attack on 76-Cr, or log ko = –1.53 for cyclization of 79-Cr versus log ko = 1.19 for cyclization of 78-Cr. These findings are reminiscent of the lower intrinsic rate constants for thiolate ion addition to vinylic substrates with a MeS leaving group compared to those with a MeO leaving group and hence must have similar explanations in terms of inductive, steric and possibly p-donor effects. Specifically, the stronger inductive effect of the MeO(RO) group enhances ko(OR) relative to ko(SR) while the larger steric effect of the MeS(RS) group lowers ko(SR) relative to ko(OR); both factors lower the ko(SR)/ko(OR) ratios. As discussed for the SNV reactions, the p-donor effects may partially offset the inductive and steric effects because early loss of the resonance stabilization of the carbene complex should lower ko(OR) more than ko(SR). However, based on our discussion of the p-donor effects in the deprotonation of Fischer carbene complexes, the ko-increasing preorganization effect may counteract or even override the ko-reducing effect of the early loss of carbene complex resonance and hence contribute to the lower ko(SR)/ko(OR) ratios.

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

307

3. The intrinsic rate constants for thiolate ion addition to 76-Cr and 76-W are substantially larger than those for alkoxide ion addition. This is similar to the previously mentioned higher intrinsic reactivity of thiolate ions compared to amine nucleophiles for the addition to a-nitrostilbene and b-nitrostyrene. It can be understood in terms of the soft–soft interaction of the thiolate ion with the carbene complex which is more advanced than C–S bond formation at the transition state.184 Transition state imbalances The fact that the intrinsic rate constants for nucleophilic addition to Fischer carbene complexes are relatively low, for example, much lower than for most reactions with comparable vinylic substrates or carboxylic esters,188 constitutes strong evidence for the presence of substantial transition state imbalances. However, there have only been a few studies of substituent effects that demonstrate the imbalance directly by showing annuc > bnnuc or by providing an estimate of its magnitude from the difference annuc – bnnuc . One such study is the reactions of 76-Cr-Z and 76-W-Z with HC CCH2O and CF3CH2O.183 It yielded annuc = 0.59 and bnnuc £ 0.46 for 76-Cr-Z, and annuc = 0.56 and bnnuc £ 0.42 for 76-W-Z, i.e., annuc > bnnuc as expected.

OMe (CO)5M

C

Z 76-Cr-Z (M = Cr) 76-W-Z (M = W)

Desolvation of the nucleophile There exists substantial evidence that in reactions that involve oxyanions or amines as bases or as nucleophiles, their partial desolvation, as they enter the transition state, typically has made greater progress than bond formation. In the context of the PNS, this partial loss of solvation represents the early loss of a reactant stabilizing factor and hence reduces the intrinsic rate constant. As discussed at some length in our 1992 chapter,4 for strongly basic oxyanions this desolvation effect often manifests itself in terms of negative deviations from Br½nsted plots and/or in abnormally low b or bnuc values.58,188 In fact, a number of cases have been reported where the bnuc value was close to zero or

308

C.F. BERNASCONI

even negative. Examples of negative bnuc values include the reaction of quinuclidines with aryl phosphates,193 of amines with carbocations,194,195 and of oximate ions with electrophilic phosphorous centers.192,196,197 The reactions of thiolate ions with several carbene complexes are also characterized by substantially negative bnuc values: they are –0.28 for 76-Cr,184 –0.25 for 76-W,184 –0.24 for 77-Cr,186 –0.30 for 77-W,186 –0.18 for 83-Cr,198 and –0.21 for 84-Cr198; for the reaction of 84-Cr with aryloxide ions bnuc = –0.39.199

O (CO)5Cr

C

O (CO)5Cr

Ph 83-Cr

NO2

C Ph 84-Cr

According to Jencks et al.,193 negative bnuc values result from a combination of minimal progress of bond formation at the transition state and the requirement for partial desolvation of the nucleophile before it enters the transition state. In a first approximation bnuc may be expressed by Equation (50) where bd and bnuc are defined by Equations (51) and (52), respectively. Kd represents 0

bnuc ¼ bd þ b nuc

ð50Þ

bd ¼ dlogKd =dpKaNucH

ð51Þ

0

0

b nuc ¼ dlogk1 =dpKaNucH

ð52Þ

the equilibrium constant for partial desolvation of the nucleophile while k0 1 is the rate constant for nucleophilic attack by the partially desolvated nucleophile. Since desolvation becomes more difficult with increasing basicity of the nucleophile, bd < 0 which, along with a small b0 nuc value, can lead to a negative bnuc value. A more elaborate treatment of this problem has been presented elsewhere.184 The fact that b0 nuc is so small as to lead to negative bnuc values implies a very small degree of C–S or C–O bond formation at the transition state. One factor that seems to play a role is the particularly severe steric crowding in the transition state due to the very large size of the (CO)5M group.186,200 The small degree of bond formation would seem to reduce the steric repulsion. The even more negative bnuc value for the aryloxide ion reactions is probably the result of a more negative bd value due to the stronger solvation of oxyanions compared to thiolate ions.

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

309

REACTIONS INVOLVING CARBOCATIONS

Following previous studies of reactions of carbocations with nucleophiles201–204 discussed in our 1992 chapter,4 Richard’s group205,206 reports that electronwithdrawing a-substituents in 4-methoxybenzyl cations, 85, reduce the rate of nucleophilic addition of alcohols and water to

R1

+ C

R2

R1

R2 C

R1, R2 = H, CF3; CF3, CF3; H, CH2F; H, CHF2; H, COOEt; CH3, CF3

OMe 85

+ OMe 85±

these cations. This is contrary to what one would expect since the electronwithdrawing a-substituents destabilize the carbocation and should make it more reactive. The reason for the reduced reactivity is that the electronwithdrawing substituents lead to a stronger resonance effect by the methoxy group. Hence the PNS effect of the early loss of resonance stabilization at the transition state increases the intrinsic barrier sufficiently as to lower the actual rate of the reaction. A similar study by Schepp and Wirz had led to the same conclusion.207 For an interesting example where the small degree of transition state resonance stabilization corresponds to a late development of product resonance is the acid-catalyzed aromatization of benzene cis-1,2-dihydrodiols.208 The reaction is shown in Scheme 2 where the loss of water is rate limiting. The rate constants as a function of 17 Z-substituents gave a good correlation with the regular Hammett s values with r = –8.2. Interestingly, there were no deviations from the Hammett plot for the p-donor substituents MeO and EtO, i.e., there was no need to use sþ constants, implying that resonance stabilization of the transition state is of minor importance despite the strongly developed positive charge indicated by the very large r value. A situation where the late development of a product destabilizing factor lowers the intrinsic barrier is the nucleophilic addition reaction shown in Equation (53).209 Kinetic data for this reaction and the reaction of a series of thiols have led to the following conclusions. The adduct, 87, is strongly stabilized by the polar and polarizability effect of the two methyl groups on the sulfur but strongly destabilized by the electron-withdrawing CF3 groups. There is also a relatively strong stabilization of the incipient positive charge on

310

C.F. BERNASCONI

Z

Z OH

OH

OH

H+ + OH2

OH

Z

Z

OH slow –H2O

+

+

H

H

hydride shift

–H+

Z

Z OH

OH –H+

+ H H

Scheme 2 + SMe2 F3C

F3C

CF3

+ CF3

CF3

F3C

(53)

+ Me2S

O–

O–

O 86

87

the sulfur atom by the methyl groups at the transition state as indicated by bnuc > 0.5 based on the addition of thiols, but only a small destabilization of the transition state by the more distant CF3 groups. The picture that emerges is that of a transition state where bond formation to the nucleophile develops at a relatively large distance so that the interaction between the positive charge and the CF3 groups remains weak until after the transition has passed while the interaction between the positive charge and the methyl groups can be strong. This, then, leads to a lowering of the intrinsic barrier. A case where the late solvation of halogen leaving groups in a carbocation forming solvolysis reaction increases the intrinsic barrier is the one shown in Equation (54). CH2X

+

CH2

+ X–

OMe

OMe

(54)

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

311

Toteva and Richard210 showed that DG‡o for F expulsion is about 3 kcal mol1 higher than for Cl expulsion. Since solvation of the fluoride ion is much stronger than that of the chloride ion, the difference in DG‡o must arise from the PNS effect of late solvation. A PNS effect involving anomeric stabilization of a geminal dialkoxy compound has been observed when comparing kinetic and thermodynamic data of reactions 55 and 56. Reaction 55

H CF3CH2O

C

OCH2CF3

CF3CH2O

+ C

H

H+ +

+ CF3CH2OH OMe

(55)

OMe

88-(OCH3CF3)2

88-OCH2CF3

H CF3CH2S

C

OCH2CF3

CF3CH2S

+ C

H

H+ +

+ CF3CH2OH OMe

88-(OCH2CF3)(SCH2CF3)

(56)

OMe 88-SCH2CF3

was reported to be thermodynamically less favorable than reaction 56 but the rate for reaction 55 is higher than for reaction 56.211 One possible interpretation of these results offered by the authors is that stabilization of 88(OCH2CF3)2 by the geminal interaction of the two oxygens, the anomeric effect,212–216 is responsible for the less favorable thermodynamics of reaction 55 but that the loss of this interaction lags behind C–O bond cleavage at the transition state. This late loss of a reactant stabilizing factor results in a lower intrinsic barrier for reaction 55.

312

C.F. BERNASCONI

MISCELLANEOUS REACTIONS

Gas phase SNV reactions Kon ´ ˘ a et al.217 reported DFT calculations on gas phase SNV reactions such as Equations (57), (58), and other similar processes. Their calculations show the expected strong stabilization of the anionic adduct HO– + CH2

CH

– CH2

OCH3

OCH3 CH

CH2

CH

OH + CH3O–

OH

(57)

HO– + O

CH

CH

CH

– O

OCH3

OCH3 CH

CH

CH

(58)

OH O

CH

CH

CH

OH + CH3O–

in Equation (58) that results from the delocalization of the negative charge onto the carbonyl oxygen. As to the barriers, the one for the first step in Equation (58) is lower than that for the first step in Equation (57) but not by an amount that would imply a strong expression of adduct stabilization in the transition state. In other words, resonance stabilization of the transition state lags behind C–O bond formation. A similar situation exists for the second step in reaction 58, i.e., the barrier is disproportionately high because of early loss of the resonance stabilization of the intermediate. Intramolecular SN2 reactions The intramolecular SN2 reaction shown in Equation (59) is an example where the development of a product destabilizing factor lags behind bond formation which contributes to the lowering of the S

–

S CH2

CH2

CH2

CH2

+ HS–

(59)

SH

CH3S– + CH3

CH2

CH3 SH

CH2 SCH3 + HS–

(60)

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

313

intrinsic barrier. Specifically, an ab initio calculation by Gronert and Lee218 has shown that the enthalpic barrier DH‡ for Equation (59) (19 kcal mol1) is lower than for Equation (60) (24 kcal mol1), even though Equation (60) is thermodynamically much more favored (DHo = –1 kcal mol1) than Equation (59) (DHo = 19 kcal mol1). This means that the intrinsic barrier of Equation (59) is much lower than for Equation (60). The cyclization is thermodynamically unfavorable due to the large ring strain of the threemembered ring. However, at the transition state, the ring strain is small because the developing C–S bond is quite long, and hence this lowers the intrinsic barrier. Another factor that contributes to the lowering of the intrinsic barrier is what the authors call the proximity effect. This effect derives from the fact that in the cyclization reaction the nucleophilic atom is forced to be close to the a-carbon which amounts to a destabilization of the substrate by 1,3-repulsive interactions. Similar results were also reported for the reactions of Equation (61).219 X CH

– XCH CH2

CH2

CH2

(61)

CH2 + Cl–

Cl X CH O, C

CH, CN

Epoxidation of alkenes Based on a kinetic study of the epoxidation of alkenes by m-chloroperbenzoic acid, Equation (62), O

O

OH

C

C

OH

O

C

C

O

+

+ C

R

(62)

C

Cl

Cl

R

R = alkyl or aryl

Perrin’s group220 concluded that, for aromatic alkenes (R = aryl), the transition state, schematically represented as 89, may be imbalanced in that the delocalization of the positive

δ+ C

C

+ •

‡

X δ– δ+

C

Ar

Ar 89

C

90

314

C.F. BERNASCONI

charge into the aromatic ring is delayed. Specifically, they showed that the kinetic data correlated with the ionization potential of the alkenes, implying that the radical cation 90 with a significant fraction of the positive charge delocalized into the aryl group, may serve as a model for the product. There were two separate correlation lines, one for aliphatic (R = alkyl) and the other for aromatic alkenes (R = aryl), and, for a given ionization potential, the reactivity of the aromatic alkenes was lower than that of their aliphatic counterparts. These results were interpreted as being the consequence of the above-mentioned transition state imbalance, which raises the intrinsic barrier of the reaction and explains the lower reactivity of the aromatic alkenes as well as the lower sensitivity of the rate to the ionization potential. Hemiacetal decomposition McClelland et al.221 have suggested that the general acid-catalyzed decomposition of a hemiacetal anion, Equation (63), proceeds through an imbalanced transition state where sp3 to sp2

‡ O–

Ar

+ HA

C CH3

OR

Oδ–

Ar

ArCCH3 + ROH + A–

C CH3

O

OR

(63)

H Aδ–

rehybridization of the central carbon lags behind C–O bond cleavage. This imbalance is in the expected direction since sp2 hybridization allows development of the acetophenone resonance. The authors based their conclusion on a relatively large blg value (large degree of C–O bond cleavage) and a small r value determined from the aryl substituent effect (small degree of charge development in the aryl group). A similar conclusion has been reached by Kandanarachchi and Sinnott222 for the hydrolysis of orthocarbonates such as (ArO)4C, (ArO)2C(OAr0 )2, or (ArO)3COAr0 . The rate-limiting step in these reactions is the spontaneous or general acid-catalyzed cleavage of the bond between the central carbon and the oxygen of the least basic aryloxy group. Again, r and blg values suggest that resonance development in the resulting carbocation lags behind C–O bond cleavage.

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

315

Radical reactions In an attempt to understand the origin of the barrier in the fragmentation of the radical cations of 2-substituted benzothiazoline derivatives, Shukla et al.223 examined the kinetics of reaction 64 as •+ S

R

S

N

Me

+ N

Me

Me + R•

(64)

Me

a function of R (PhCH2, Ph2CH, PhCMe2, 9-MeFl, Ph2CMe) and performed DFT energy calculations. Even though the activation barriers decreased with increasing resonance stabilization of the radical R, it was shown that the lowering of the barriers was rather modest relative to the large increases in the thermodynamic driving force that resulted from the enhanced radical stability. This unusually small effect on the barrier was attributed to an increase in the intrinsic barrier with increasing resonance stabilization of the radical. The increase in the intrinsic barrier was attributed to ‘‘large reorganization energies in the product fragments.’’ The authors did not refer to the transition state as ‘‘imbalanced’’ and apparently were unfamiliar with the PNS, but their results are of course a classic example of the PNS at work. In a theoretical paper, Costentin and SavO˜ant224 examined the dimerization of neutral radicals by constructing potential energy profiles from AMI calculations, subjecting some selected dimerizations to B3LYP/6-31G* calculations and applying VB theory in analyzing the results. The dimerization of conjugated radicals, e.g., Equation (65), is subject to an 2 CH2

CH

CH2

CH2

CH

CH2

CH2 CH

CH2

(65)

activation barrier which contrasts with the barrierless dimerization of nonconjugated radicals; these barriers are higher for more highly delocalized radicals, even when the reactions are more thermodynamically favored. These results are consistent with the notion that loss of the resonance of the radicals is ahead of bond formation, or, in the reverse direction, development of the radical resonance lags behind bond cleavage. However, based on their analysis of the reaction in the reverse direction for which the resonance integral apparently increases at the same pace as bond cleavage, Costentin and SavO˜ant concluded that delocalization is synchronous with bond breaking. It would appear that further study is needed to resolve this apparent inconsistency between their conclusion and the predictions of the PNS.

316

C.F. BERNASCONI

Enzyme-catalyzed hydride transfer A biologically relevant example of a reaction with an imbalanced transition state is the hydride transfer catalyzed by dihydrofolate reductase of Escherichia coli. In a theoretical study by Pu et al.225 the hybridization changes at the donor carbon atom (C4N) and acceptor carbon atom (C6) that occur along the reaction coordinate were examined. It was shown that the changes in hybridization at both carbon centers progress in a nonlinear fashion with respect to the progress of the hydride transfer. Specifically, the change from sp3 to sp2 hybridization of C4N lags behind hydride transfer while the change from sp2 to sp3 hybridization of C6 is ahead of hydride transfer. This is strictly analogous to the findings for the proton transfer reactions of the type of Equation (21) where the change in the pyramidal angle (56) lags behind proton transfer. Additional evidence for the imbalanced nature of the transition state was deduced from an analysis of the changes in the C4–H and C6–H bond orders along the reaction coordinate.

5

Summary and concluding remarks

Most elementary reactions involve several molecular events such as bond formation/cleavage, charge transfer, charge creation/destruction, charge delocalization/localization, creation/destruction of aromaticity or antiaromaticity, increase/decrease in steric strain, etc. It is rare that all these events have made equal progress at the transition state; in other words, in most cases, the transition state is imbalanced in the sense that some process develops ahead of or lags behind others along the reaction coordinate. It has proven useful to regard the main bond changes as the ‘‘primary’’ process and to regard the development of the various product stabilizing/destabilizing factors, or the loss of the various reactant stabilizing/destabilizing factors, as ‘‘secondary’’ processes. This definition then allows us to use the extent of the bond changes as a frame of reference in gauging whether the development of a product stabilizing/destabilizing factor or the loss of a reactant stabilizing/destabilizing factor is early or late. Within this framework the various manifestations of the PNS summarized in Chart 1 are unambiguous and there can be no exceptions. What makes the PNS universal is that it is applicable to all reactions that involve bond changes. It provides a qualitative and sometimes even semiquantitative understanding of chemical reactivity using the language of physical organic chemistry. Its main virtue and usefulness is that, for the most part, a given factor follows a consistent pattern, i.e., it invariably either develops late or early, regardless of the specific reactions, and hence its effect on the intrinsic barrier is predictable. A summary of how the various factors discussed in this chapter affect intrinsic barriers/intrinsic rate constants is provided in Table 26. They include charge delocalization/resonance, solvation, aromaticity, anti-

Factor

1

Effect on molecule

Late development/ early loss

Early development/ late loss

Effect on DG‡o

ko

"

#

Ubiquitous, no exceptions, predicted by theory [Equations (12) and (25)]

" #

# "

Ubiquitous, no exceptions Limited number of known cases [Equations (18–20), (33a), (34a), and (36)] One established case [Equation (35a)], one tentative case (deprotonation of 51 and 52) Numerous cases, e.g., deprotonation of HOCH2CH2NO2, PhCH2CH2NO2, (CH3)2CHNO2, and Fischer carbenes Effect sometimes masked by preorganization (deprotonation of Fischer carbenes, nucleophilic additions) Limited number of cases (nucleophilic additions) Limited number of cases (RS as nucleophile) One tentative case [Equation (55)] Limited number of cases, e.g., 65 [Equation (45)]

Stabilizing

H

2 3

Charge delocalization (resonance) Solvation Aromaticity

Stabilizing Stabilizing

H

4

Anti-aromaticity

Destabilizing

H

#

"

5

Hyperconjugation

Stabilizing

H

"

#

6

p-Donor effect

Stabilizing

H

"

#

7

Steric effect (Fstrain) Soft–soft interactions Anomeric effect Intramolecular hydrogen bonding Ring strain

Destabilizing

H

"

#

Stabilizing

H

#

"

H

8 9 10 11

H

Comments

Stabilizing Stabilizing

H

# "

" #

Destabilizing

H

#

"

317

Limited number of cases, e.g., Equations (59) and (61)

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

Table 26 The effect of product/reactant stabilizing/destabilizing factors on DG‡o and ko

318

C.F. BERNASCONI

aromaticity, hyperconjugation, p-donor effects, F-strain, soft–soft interactions, and anomeric effects. There are other factors that can affect DG‡o and ko; they do so not because they inherently develop nonsynchronously relative to bond changes but because of delayed charge delocalization. They include the effects of remote as well as adjacent polar and polarizable substituents, of remote and adjacent charge, and of substituents that impede charge delocalization by steric crowding. They are summarized in Table 27. What the PNS cannot deal with is the effect on reactivity by factors that only operate at the transition state level but are not present in either reactant or product. Examples mentioned in this chapter include transition state aromaticity in Diels Alder reactions, steric effects on reactions of the type A þ B ! C þ D, or hydrogen bonding/electrostatic effects that stabilize the

Table 27 The effect of substituents and charges on DG‡o and ko for reactions with imbalanced transition states Factor

1 2 3

4

5 6

EW polar substituent close to charge of TS ED polar substituent close to charge of TS EW polar substituent far from charge at TS ED polar substituent far from charge at TS Adjacent polarizable substituent Adjacent positive charge

7

Remote positive charge

8

Steric hindrance of resonance by substituent

Effect on TS versus effect on product

Effect on

Comments

DG‡o

ko

#

"

"

#

"

#

Disproportionately small TS destabilization

#

"

Equation (14) (leads to aCH < bB); Equation (6)

Disproportionately large TS stabilization Disproportionately large TS stabilization Disproportionately small TS stabilization Disproportionately small TS destabilization

#

"

#

"

Proton transfer progress must be substantial e.g., Equation (16)

"

#

e.g., nucleophilic addition to 85

#

"

e.g., CH3CH(NO)2 versus CH3NO2 deprotonation

Disproportionately large TS stabilization Disproportionately large TS destabilization Disproportionately small TS stabilization

Equation (4) (leads to aCH > bB); Equation (15) Equation (4) (leads to to aCH > bB); Equation (15) Equation (14) (leads to aCH < bB); Equation (6)

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

319

transition state of proton transfers of the type CH3Y þ CH2=Y ! CH2Y=Y þ CH3Y which are especially strong in the gas phase.

Acknowledgments I gratefully acknowledge the many outstanding contributions of all of my coworkers whose names are cited in the references and the financial support by the National Science Foundation (grant no. CHE-0446622).

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29.

Bernasconi CF. Tetrahedron 1985;41:3219. Bernasconi CF. Acc Chem Res 1987;20:301. Bernasconi CF. Acc Chem Res 1992;25:9. Bernasconi CF. Adv Phys Org Chem 1992;27:119. Marcus RA. J Chem Phys 1956;24:966. Marcus RA. J Chem Phys 1957;26:867. Marcus RA. J Phys Chem 1968;72:891. Keeffe JR, Kresge AJ. In: Bernasconi CF, editor. Investigation of rates and mechanisms of reactions, Part 1. New York: Wiley-Interscience; 1986, p. 747. For proton transfers statistical factors p for the number of protons in the buffer acid and q for the number of equivalent basic sites on the buffer base are typically applied.8 Jencks DA, Jencks WP. J Am Chem Soc 1977;99:451. Jencks WP. Chem Rev 1985;85:54. Bunnett JF. Angew Chem Int Ed Engl 1962;1:225. More O’Ferrall RA. J Chem Soc (B) 1970:274. Saunders WH, Jr, Cockerill AF. 1973. Mechanisms of elimination reactions. New York: Wiley. Eigen M. Angew Chem. Int Ed Engl 1964;3:1. Ahrens M-L, Maass G. Angew Chem Int Ed Engl 1968;7:818. Crooks JE. In: Caldin EF, Gold V, editors. Proton transfer reactions. New York: Wiley; 1975. p. 153. Hibbert F. Compr Chem Kin 1977;8:97. Bernasconi CF, Terrier F. J Am Chem Soc 1987;108:7115. Bernasconi CF, Fairchild DE, Montaez RL, Aleshi P, Zheng H, Lorance E. J Org Chem 2005;70:7121. Bernasconi CF, Hibdon SA. J Am Chem Soc 1983;105:4343. Bernasconi CF, Sun W. J Am Chem Soc 1993;115:12526. Bernasconi CF, Paschalis P. J Am Chem Soc 1986;108:2969. Bernasconi CF, Bunnell RD. Isr J Chem 1985;26:420. Bernasconi CF, Kliner DAV, Mullin AS, Ni JX. J Org Chem.1988;53:3342. Moutiers G, Peigneux A, Vichard D, Terrier F. Organometallics 1998;17:4469. Terrier F, Xie H-Q, Farrell PG. J Org Chem 1990;55:2610. Terrier F, Croisat D, Chatrousse A-P, Pouet M-J, HallO˜ J-C, Jacob G. J Org Chem 1992;57:3684. Leffler JE, Grunwald E. Rates and equilibria in organic reactions. New York: Wiley; 1963. p. 156.

320

C.F. BERNASCONI

30. Kresge AJ. In: Caldin EF, Gold V, editors. Proton transfer reactions. New York: Wiley; 1975. p. 519. 31. Bordwell FG, Boyle WJ, Jr. J Am Chem Soc 1972;94:3907. 32. Bowden K, Hirani SIJ. J Chem Soc, Perkin Trans 2 1990:1889. 33. Bell RP, Grainger S. J Chem Soc, Perkin Trans 2 1976:1367. 34. Yao X, Gold MA, Pollack RM. J Am Chem Soc 1999;121:6220. 35. Gandler JR, Bernasconi CF. J Am Chem Soc 1992;114:631. 36. Keeffe JR, Morey J, Palmer CA, Lee JC. J Am Chem Soc 1979;101:1295. 37. Terrier F, LeliO`vre J, Chatrousse A-P, Ferrall P. J Chem Soc, Perkin Trans 2 1985:p1479. 38. Gandler JR, Saunders OL, Barbosa R. J Org Chem 1997;62:4677. 39. Barletta GL, Zou Y, Huskey WP, Jordan F. J Am Chem Soc 1997;119:2356. 40. Bernasconi CF, Wenzel PJ. J Am Chem Soc 1996;18:11446. 41. Bernasconi CF, Fairchild DE. J Phys Org Chem 1992;5:409. 42. Murray CJ, Jencks WP. J Am Chem Soc 1990;112:1880. 43. Stefanidis D, Bunting JW. J Am Chem Soc 1991;113:991. 44. Wodzinksi S, Bunting JW. J Am Chem Soc 1994;116:6910. 45. Fogel P, Farrell PG, LeliO`vre J, Chatrousse A-P, Terrier F. J Chem Soc, Perkin Trans 2 1985:711. 46. Nevy JB, Hawkinson DC, Blotny G, Pollack RM. J Am Chem Soc 1997;119:12722. 47. Alston WC, II, Haley K, Kauski R, Murray CJ, Pramata J. J Am Chem Soc 1996;118:6562. 48. Zhong Z, Snowden TS, Best MD, Anslyn EV. J Am Chem Soc 2004;126:3488. 49. Amyes TL, Richard JP. J Am Chem Soc 1996;118:3129. 50. Angelini G, DeMaria P, Fontana A, Pierini M, Siani G. J Org Chem 2007;72:4039. 51. Kresge AJ. Can J Chem 1974;52:1897. 52. Pross A, Shaik SS. J Am Chem Soc 1982;104:1129. 53. Pross A. Adv Phys Org Chem 1985;21:99. 54. Bernasconi CF, Kittredge KW. J Org Chem 1998;63:1944. 55. Bernasconi CF, Montaez RL. J Org Chem 1997;62:8162. 56. Bernasconi CF, Bunnell RD, Terrier F. J Am Chem Soc 1988;110:6514. 57. Bernasconi CF, Bunnell RD. J Am Chem Soc 1988;110:2900. 58. There is substantial evidence that solvation of oxyanions is ahead of proton transfer as well as ahead of C–O bond formation in nucleophilic reactions.42,59–65 59. Kresge AJ. Chem Soc Rev 1973;2:475. 60. Hupe SJ, Jencks WP. J Am Chem Soc 1977;99:451. 61. Pohl ER, Wu D, Hupe DJ. J Am Chem Soc 1980;102:2759. 62. Jencks WP, Brant SR, Gandler JR, Fendrich G, Nakamura C. J Am Chem Soc 1982;108:479. 63. Terrier F, Degorre F, Kiffer D, Laloi M. Bull Soc Chim Fr 1988:415. 64. Terrier F, Moutier G, Xiao L, Le GuO˜vel E, Guir F. J Org Chem 1995;60:1748. 65. Buncel E, Cannes C, Chatrousse A-P, Terrier F. J Am Chem Soc 2002;124:8666. 66. Bernasconi CF, PO˜rez-Lorenzo M, Brown SD. J Org Chem 2007;72:4416. 67. Bernasconi CF, Kanavarioti A. J Org Chem 1979;44:4829. 68. Bernasconi CF, Panda M, Stronach MW. J Am Chem Soc 1995;117:9206. 69. Terrier F, Kizilian E, Goumont R, Faucher N, Wakselman C. J Am Chem Soc 1998;120:9496. 70. Terrier F, Magnier E, Kizilian E, Wakselman C Buncel E. J Am Chem Soc 2005;127:5563. 71. Hansch C, Leo A, Taft RW. Chem Rev 1991;91:165.

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

321

72. Bernasconi CF, Sun W, GarcU´a-RU´o L, Yan K, Kittredge KW. J Am Chem Soc 1997;119:5582. 73. Bernasconi CF, Leyes AE, GarcU´a-RU´o L. Organometallics 1998;17:4940. 74. McMurry J. Organic chemistry, 7th ed. Belmont, California: Thomson-Brooks/ Cole; 2008. pp. 187. 75. Goumont R, Magnier E, Kizilian E, Terrier F. J Org Chem 2003;68:6566. 76. Goumont R, Kizilian E, Buncel E, Terrier F. Org Biomol Chem 2003;1:1741. 77. Price CC, Oae S. Sulfur bonding. New York: Ronald Press; 1962. 78. Oae S, Tagaki W, Ohno A. Tetrahedron 1964;20:417. 79. Eliel EL, Hartmann AA, Abatjoglon AG. J Am Chem Soc 1974;96:1807. 80. Wolfe S, La John LA, Bernardi F, Mangini A, Tonachini G. Tetrahedron Lett 1983;24:3789. 81. Bernardi F, Mangini A, Tonachini G, Vivarelli P. J Chem Soc, Perkin Trans 2 1985:111. 82. Bordwell FG, Bares JE, Bartmess JE, Drucker GF, Gerhold J, McCollum JG, et al. J Org Chem 1977;42:326. 83. Streitweiser A, Jr, William JE. J Am Chem Soc 1975;97:191. 84. Lehn J-M, Wipf G. J Am Chem Soc 1976;98:7498. 85. Bernardi F, Csizmadia IG, Mangini A, Schlegel HB, Wangbo M-H, Wolfe S. J Am Chem Soc 1975;97:2209. 86. Schleyer PvR, Clark T, Kos AJ, Spitznagel G, Rohde C, Arad D, et al. J Am Chem Soc 1984;106:6467. 87. Hopkinson AC, Lien MH. J Org Chem 1981;46:998. 88. Schleyer PvR, Kos AJ. Tetrahedron 1983;39:1141. 89. Wiberg KB, Castejon H. J Am Chem Soc 1994;116:10489. 90. Cuevas G, Juaristi E. J Am Chem Soc 1997;119:7545. 91. Taft RW, Topsom RD. Prog Phys Org Chem 1987;119:7545. 92. Keeffe JR, Kresge AJ. In: Rappoport Z, editor. The chemistry of enols. New York: Wiley & Son; 1990. p. 399. 93. Carey ARE, Al-Quatomi S, More O’Ferrall RA, Murray BA. J Chem Soc, Chem Commun 1988:1097. 94. Bernasconi CF, Moreira JA, Huang LL, Kittredge KW. J Am Chem Soc 1999;121:1674. 95. Tobin JB, Frey PA. J Am Chem Soc 1996;118:12250. 96. Bernasconi CF, Sun W. Organometallics 1997;16:1926. 97. Bernasconi CF, Leyes AE. J Am Chem Soc 1997;117:5169. 98. Bernasconi CF, Ali M. J Am Chem Soc 1999;121:3039. 99. Hafner A, Hegedus LS, de Weck G, D—tz KH. J Am Chem Soc 1988;110:8413. 100. Bernasconi CF, Leyes AE, Ragains ML, Shi Y, Wang H, Wulff WD. J Am Chem Soc 1998;120:8632. 101. Bernasconi CF, Sun W. J Am Chem Soc 2002;124:2299. 102. Bernasconi CF, Ruddat V. J Am Chem Soc 2002;124:14968. 103. Bernasconi CF, Ali M, Gunter JC. J Am Chem Soc 2003;125:151. 104. Bernasconi CF, Ragains ML, Bhattacharya S. J Am Chem Soc 2003;125:12328. 105. The aromaticity of the heterocycles follows the order thiophene > selenophene > furan.106,107 106. Bird CW. Tetrahedron 1985;41:1409. 107. Minkin VI, Glukhoutsev MN, Simkin BY. Aromaticity and anti-aromaticity. New York: Wiley-Interscience; 1994. 108. Bernasconi CF, PO˜rez-Lorenzo M. J Am Chem Soc 2007;129:2704. 109. Bernasconi CF, Zheng H. J Org Chem 2006;71:8203. 110. Bernasconi CF, Ruddat V, Wenzel PJ, Fischer H. J Org Chem 2004;69:5232.

322

C.F. BERNASCONI

111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121.

Dega-Szafran Z, Schroeder G, Szafran M. J Phys Org Chem 1999;12:39. Saunders WJ, Jr. J Am Chem Soc 1994;116:5400. Bernasconi CF, Wenzel PJ. J Am Chem Soc 1994;116:5405. Saunders WH, Jr, Van Verth JE. J Org Chem 1995;60:3452. Harris N, Wei W, Saunders WH, Jr, Shaik S. J Am Chem Soc 2000;122:6754. Lee I, Kim CK, Kim CK. J Phys Org Chem 1999;12:255. Bernasconi CF, Wenzel PJ, Keeffe JR, Gronert S. J Am Chem Soc 1997;119:4008. Bernasconi CF, Wenzel PJ. J Org Chem 2001;66:968. Harris N, Wei W, Saunders WH, Jr, Shaik S. J Phys Org Chem 1999;12:259. Bernasconi CF, Wenzel PJ. J Am Chem Soc 1996;118:10494. Possible reasons for the exceptionally large n values for CH3C CH have been discussed elsewhere.118 Yamataka H, Mustanir, Mishima M. J Am Chem Soc 1999;121:10225. Sato M, Kitamura Y, Yoshimura N, Yamataka H. J Org Chem 2009;74:1268. More O’Ferrall RA. J Chem Soc B 1970:274. Jencks WP. Chem Rev 1972;72:705. Costentin C, SavO˜ant J-M. J Am Chem Soc 2004;126:14787. In solution there is also solvational reorganization.126 Taft RW, Topsom RD. Prog Phys Org Chem 1987;16:1. Taft RW. Prog Phys Org Chem 1983;14:247. Van Verth JE, Saunders WH, Jr. J Org Chem 1997;62:5743. Farneth WE, Brauman JI. J Am Chem Soc 1976;98:7891. Pellerite MJ, Brauman JI. J Am Chem Soc 1980;102:5993. Moylan CR, Brauman JI. Annu Rev Phys Chem 1983;34:187. Boys SF, Bernardi F. Mol Phys 1970;19:55. Scheiner’s136–138 work is also relevant here. Cybulski SM, Scheiner S. J Am Chem Soc 1987;109:4199. Scheiner S, Wang L. J Am Chem Soc 1992;114:3650. Scheiner S. J Mol Struct (TEOCHEM) 1994;307:65. Gronert S. J Am Chem Soc 1993;115:10258. Bernasconi CF, Wenzel PJ. J Org Chem 2003;68:6870. Bernasconi CF, Wenzel PJ, Ragains. ML. J Am Chem Soc 2008;130:4934. HOMA = Harmonic Oscillator Model.139 Krygowski TM, CyrA˜nski MK. Chem Rev 2001;101:1385. NICS = Nuclear-independent chemical shift.145 Chen Z, Wannese CS, Corminboeuf C, Puchta R, Schleyer PvR. Chem Rev 2005;105:3842. The strongly negative NICS(1) value for C4 Hþ 5 has been attributed to its homoaromaticity107 which is consistent with the non-planar geometry of C4 Hþ 5 . The strongly negative NICS(1) value for the transition state suggests that the homo141 aromaticity of C4 Hþ 5 is highly preserved at the transition state. Bernasconi CF, Yamataka H, Yoshimura N, Sato M. J Org Chem 2009;74:188. Bird CW. Tetrahedron 1985;41:1409. Bird CW. Tetrahedron 1992;48:335. Evans MG. Trans Faraday Soc 1939;35:824. Dewar MJS. The molecular orbital theory of organic chemistry. New York: McGraw-Hill; 1969. pp. 316–9. Zimmerman H. Acc Chem Res 1971;4:272. Herges R, Jiao H, Schleyer PvR. Angew Chem Int Ed Engl 1994;33:1376. CossU´o FP, Morao I, Jiao H, Schleyer PvR. J Am Chem Soc 1999;121:6737. Corminboeuf C, Heine T, Weber J. Org Lett 2003;5:1127.

122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146.

147. 148. 149. 150. 151. 152. 153. 154. 155.

THE PRINCIPLE OF NONPERFECT SYNCHRONIZATION

323

156. Stahl F, Moran D, Schleyer PvR, Prall M, Schreiner PR. J Org Chem 2002;67:1453. 157. Fernandez I, Sierra MA, CossU´o FP. J Org Chem 2007;72:1488. 158. Schleyer PvR, Maerker C, Dransfeld A, Jiao A, van Eikema Hommes WJR. J Am Chem Soc 1996;118:6317. 159. Tsuij T, Okuyama M, Ohkita M, Kawai H, Suzuki T. J Am Chem Soc 2003;125:951. 160. PhCH=C(C4Cl4) = 1,2,3,4-tetrachloro-6-phenylfulvene; C5H2Cl4 = 1,2,3,4-tetrachloro-1,3-cyclopentadiene. 161. Atherton JH, Crampton MR, Duffield GL, Steven JA. J Chem Soc, Perkin Trans 2 1995:443. 162. Cox JP, Crampton MR, Wight P. J Chem Soc, Perkin Trans 2 1988:25. 163. Rappoport Z. Adv Phys Org Chem 1969;7:1. 164. Modana G. Acc Chem Res 1971;4:73. 165. Rappoport Z. Acc Chem Res 1981;14:7. 166. Rappoport Z. Recl Trav Chim Pays-Bas 1988;104:309. 167. Shainyan BA. Russ Chem Rev 1986;55:511. 168. Rappoport Z. Acc Chem Res 1992;25:474. 169. Okyama T, Lodder G. Adv Phys Org Chem 2002;37:1. 170. Bernasconi CF, Rappoport Z. Acc Chem Res 2009:42:993. 171. Bernasconi CF, Ketner RJ, Ragains ML, Chen X, Rappoport Z. J Am Chem Soc 2001;123:2155. 172. Bernasconi CF, Killion RB, Jr. J Am Chem Soc 1988;110:7506. 173. Bernasconi CF, Ketner RJ. J Org Chem 1998;68:6266. 174. Bernasconi CF, Fassberg J, Killion RB, Jr., Rappoport Z. J Am Chem Soc 1990;112:3169. 175. Bernasconi CF, Ketner RJ, Chen X, Rappoport Z. J Am Chem Soc 1998;120:7461. 176. Bernasconi CF, Ketner RJ, Chen X, Rappoport Z. Can J Chem 1999;77:584. 177. MA = Meldrum’s acid, ID = indandione. 178. Bernasconi CF, Ketner RJ, Chen X, Rappoport Z. ARKIVOC 2002 (xii) 2002:161. 179. The physical organic chemistry of Fischer carbene complexes has been reviewed;176,177 only studies relevant to the PNS will be discussed in this chapter. 180. Bernasconi CF. Chem Soc Rev 1997;26:299. 181. Bernasconi CF. Adv Phys Org Chem 2002;37:137. 182. Bernasconi CF, Flores FX, Gandler JR, Leyes AE. Organometallics 1994;13:2186. 183. Bernasconi CF, GarcU´a-RU´o L. J Am Chem Soc 2000;122:3821. 184. Bernasconi CF, Kittredge KW, Flores FX. J Am Chem Soc 1999;121:6630. 185. Bernasconi CF, Ali M, Lu F. J Am Chem Soc 2000;122:1352. 186. Bernasconi CF, Ali M. J Am Chem Soc 1999;121:11384. 187. D—tz KH, Fischer H, Hofmann R, Kreissl FR, Schubert U, Weiss K. Transition metal carbene complexes. Deerfield Beach, FL: Verlag Chemie; 1983. 188. Numerous authors189–191 have compared the reactions of Fischer carbene complexes with nucleophiles to the corresponding reactions of carboxylic esters.183,185– 187 Our view is that there is much more resemblance between the reactions of Fischer carbene complexes and SNV reaction than between the reactions of Fischer carbene complexes and reactions with esters because in the latter reactions there are no strong resonance effects. 189. Schubert U, editor. Advances in metal carbene chemistry. Dordrecht, Holland: Kluwer; 1989. 190. Werner H, Fischer EO, Heckl B, Kreiter CG. J Organomet Chem 1971;28:367.

324

C.F. BERNASCONI

191. Aumann R, Hinterding P, Krger C, Goddard R. J Organomet Chem 1993;459:145. 192. Takka RM, Buncel E. J Am Chem Soc 1995;117:1503. 193. Jencks WP, Haber MT, Herschlag D, Nazaretian KL. J Am Chem Soc 1986;108:479. 194. Richard JP. J Chem Soc, Chem Commun 1987:1768. 195. McClelland RA, Kanagasabapthy VH, Banait NS, Steenken S. J Am Chem Soc 1992;114:1816. 196. Terrier F, Le GuO˜vel E, Chatrousse A-P, Moutiers G, Buncel E. Chem Commun 2003:600. 197. Terrier F, Rodriguez-Dafonte P, Le GuO˜vel E, Moutiers G. Org Biomol Chem 2006;4:4352. 198. Bernasconi CF, PO˜rez-Lorenzo M, Codding SJ. J Org Chem 2007;72:9456. 199. Bernasconi CF, Zoloff Michoff ME, deRossi RH, Granados AM. J Org Chem 2007;72:1285. 200. Bernasconi CF, Bhattacharya S. Organometallics 2003;22:426. 201. Richard JP. J Am Chem Soc 1986;108:6819. 202. Richard JP. J Am Chem Soc 1989;111:6795. 203. Richard JP, Amyes TL, Stubblefield V. J Am Chem Soc 1990;112:9513. 204. Richard JP, Amyes TL, Stevens IW. Tetrahedron Lett 1991;32:4255. 205. Amyes TL, Stevens IM, Richard JP. J Org Chem 1993;58:6057. 206. Richard JP. J Org Chem 1994;59:25. 207. Schepp NP, Wirz J. J Am Chem Soc 1994;116:11749. 208. Boyd DR, Blacker J, Byrne B, Dalton H, Hand MV, Kelly SC, et al. J Chem Soc, Chem Commun 1994:313. 209. Toteva MM, Richard JP. J Am Chem Soc 2000;122:11073. 210. Toteva MM, Richard JP. J Am Chem Soc 2002;124:9798. 211. Jagannadham V, Amyes TL, Richard JP. J Am Chem Soc 1993;115:8465. 212. Hine J. Structural effects on equilibria in organic chemistry. New York: WileyInterscience; 1975. pp. 8–10. 213. Kirby AJ. The anomeric effect and related stereoelectronic effects in organic chemistry. Oxford:Pergamon; 1983. 214. More O’Ferrall RA, Murray BA. J Chem Soc, Chem Commun 1988:1098. 215. Schleyer PvR, Jemmis ED, Sptiznagel GW. J Am Chem Soc 1985;107:6393. 216. Reed AE, Schleyer PvR, J Am Chem Soc 1987;109:7362. 217. Kon ´ ˘ a J, Zahradnik P, Fabian WMF. J Org Chem 2001;66:4998. 218. Gronert S, Lee JM. J Org Chem 1995;60:6731. 219. Gronert S, Azizian K, Friedman MA. J Am Chem Soc 1993;120:3220. 220. Kim C, Traylor TG, Perrin CL. J Am Chem Soc 1998;120:9513. 221. McClelland RA, Engelt KM, Larsen TS, S½rensen PE. J Chem Soc, Perkin Trans 2 1994:2199. 222. Kandanarachichi P, Sinnott ML. J Am Chem Soc 1994;116:5592. 223. Shukla D, Liu G, Dinnocenzo JP, Farid S. Can J Chem 2003;81:744. 224. Costentin C, SavO˜ant J-M. J Phys Chem A 2005;109:4125. 225. Pu J, Ma S, Garcia-Viloca M, Gao J, Truhlar DG, Kohen A. J Am Chem Soc 2005;127:14879.