Oswald S. Tee

The Stabilization of Transition States by Cyclodextrins and other Catalysts

Publisher Summary This chapter describes the stabilization of transition states by cyclodextrins and other catalysts. It deals with a particular aspect of the chemistry of cyclodextrins: the effects that they can have on organic reactions by virtue of their abilities to bind to many organic and inorganic species. The main emphasis of this chapter is on catalysis, because this is of greater interest, the Kurz method can also be applied to retardation. In fact, the smooth transition from retardation, through inactivity, to full catalysis can be quantified and analyzed in relation to the structure of the species concerned. Moreover, there are some examples involving catalysis by acids and bases, metal ions, micelles, amylose, catalytic antibodies, and enzymes to give an idea about the ways Kurzs approach may be usefully applied to other catalysts. This chapter concludes that the treatment addressed in the chapter stimulates use and exploration of the Kurz approach to analyzing transition state stabilization.

The binding of transition states by Cyclomalto-oligosaccharides

Abstract A method of estimating the strength of binding of transition states to catalysts is applied to reactions mediated by cyclomalto-oligosaccharides (cyclodextrins, CDs). The method affords values of the apparent dissociation constant ( K TS ) of the transition state of the “catalyzed” reaction into that of the “uncatalyzed” reaction and the “catalyst”, a CD. Variations in K TS with changes in structure are useful in the study of reactions influenced by CDs, and, in particular instances, the values of log K TS exhibit linear free energy relationships. Examples are presented where the sensitivity (or insensitivity) of the value of K TS (and log K TS ) to changes in substrate structure allows a distinction to be made between different modes of binding of the transition state.

The binding of alkyl chains to β-cyclodextrin and ‘hydroxypropyl-β-cyclodextrin’

The strength of binding of aliphatic alcohols, alkanesulfonate ions and aryl alkanoates to ‘hydroxypropyl-β-cyclodextrin’ is very similar to that to unmodified β-cyclodextrin (β-CD). The clear inference is that both forms of β-CD bind these guests to the wider opening of the CD cavity that is surrounded by secondary hydroxy groups.

Binding of aliphatic ketones to cyclodextrins in aqueous solution

Dissociation constants (Kd) for the complexation of 22 simple ketones with α-, β- and hydroxypropyl-β- cyclodextrin (α-CD, β-CD, and HP-β-CD) in aqueous solution have been determined. For these constants, there are various correlation involving pKd(=–log Kd) which have the form of linear free energy relationships. In particular, there are strong correlation between the pKd values of ketones (RCOR′) and related secondary alcohols [RCH(OH)R′], including cases where R and R′ form a ring. As with other alkyl derivatives, pKd values for alkan-2-ones and alkan-3-ones increase monotomcally with chain length, with slopes about 0.4, corresponding to Gibbs energy increments of ca. 2.3 kJ mol–1 for each CH2 group that is sequestered by the CD. The strengths of binding of linear derivatives to α-CD and β-CD correlate well, but bulky and cyclic ketones bind more weakly to α-CD, due to its smaller cavity. The pKd values for complexation of 18 of the 22 ketones by HP-β-CD and β-CD are fairly close and linearly related with a slope of 0.96 ± 0.03. These data are a subset of a larger set for 68 aliphatic compounds for which the slope is 0.99 ± 0.02. Thus, the strength of binding of such aliphatics to HP-β-CD and β-CD is generally close, although the penetration of the CD cavity by the guests is not necessarily the same for these two CDs.

Catalysis of electrophilic bromine attack by α-cyclodextrin

The aqueous bromination of various aromatic and heteroaromatic substrates is catalysed by α-cyclodextrin (α-CD), and the oxidation of formic acid (by bromine attack on formate ion) is similarly catalysed. The kinetic results are interpreted in terms of reaction between free substrate and the α-CD·Br2 complex, which is slightly more reactive than free bromine. This interpretation is consistent with the finding that substituent effects for the catalysed and uncatalysed reactions are essentially identical and that the transition state stabilization afforded by α-CD varies little for substrates having a 1010 range of reactivity.

Kinetics and mechanism of the reversible ring-opening of thiamine and related thiazolium ions in aqueous solution

Kinetic studies of the ring-opening and reclosure reactions of thiamine and three other thiazolium ions (Q+) in aqueous solution, in the pH range 0–13, have been carried out by stopped-flow and conventional UV–VIS spectrophotometry. At high pH, ring-opening of thiamine exhibits a temporary diversion to the well-known ‘yellow form’. Otherwise, the ring-opening reactions are simply first-order in [OH–], consistent with rate-limiting attack of hydroxide ion at C(2) of the Q+ ring, producing a pseudobase, T°, which rapidly consumes a second equivalent of hydroxide ion to form the ring-opened enethiolate, ETh–. In contrast, ring closure of the enethiol in acidic solution exhibits rather complex kinetic behaviour; two processes are observed for most enethiols, including that derived from thiamine. Both the fast process (a) and the slower process (b) produce the thiazolium ion Q+ and they exhibit pH- and buffer-independent rate plateaux at low pH. Rapid, repetitive UV spectral scans and NMR spectral studies show that the two processes arise from the independent formation of Q+ from the two amide rotamers of the enethiol which do not equilibrate under the reaction conditions. The major amide rotamer (∼75%) gives rise to the fast process (a) and the minor rotamer to the slow reaction (b). The pH–rate profile and buffer catalysis studies reveal that the reclosure reaction undergoes a change in rate-limiting step from uncatalysed formation of T° at low pH to its general acid catalysed breakdown at higher pH. The latter process is characterized by a Bronsted α value of 0.70. Additionally, for process (b), a general base catalysed pathway for formation of T° can be observed, for which the Bronsted β value is 0.74. The mechanistic details of the ring-opening and reclosure pathways are discussed.

Acyl transfer reactions mediated by cyclodextrins. The reaction of external nucleophiles with encapsulated alkanoate esters of varying chain length

The kinetics of the reaction of p-nitrophenyl alkanoates (acetate to decanoate, C2 to C10) with trifluoroethanol (TFE) in the presence of α-, β-, or hydroxypropyl-β-cyclodextrin (α-, β-, or Hp-β-CD) in basic aqueous solution have been measured. The results are analysed to afford rate constants for nucleophilic attack on the free and CD-bound esters (kN and kcN, respectively). Generally speaking, the values of kN and kcN are not very different, so that binding the esters to CDs has only modest effects on their reactivities towards TFE, reacting as its anion. However, there is a general trend in kcN values such that transition-state stabilization increases in a biphasic manner as the alkanoate chain is lengthened from C2 to C10. For short chains ( C6) is quite steep. This same behaviour is observed for all three CDs, with only minor differences between them, and also when the nucleophile is the anion of 2-mercaptoethanol. It is suggested that there is a change in the mode of transition-state binding of the esters from aryl group inclusion (3‡) for the short esters to acyl-group inclusion (4‡) when the acyl chain is lengthened beyond C6.

The kinetics of cleavage of nitrophenyl alkanoates by γ-cyclodextrin and by ‘dimethyl-β-cyclodextrin’ in basic aqueous solution

The cleavage of m- and p-nitrophenyl alkanoates (C2 to C8) in basic solution is accelerated modestly (7–17 times), by γ-cyclodextrin (γ-CD). The effects on the two isomeric series of esters are virtually the same and they hardly vary with the ester chain length. Cleavage of the same esters (C2 to C10) is retarded by ‘dimethyl-β-cyclodextrin’(diMe-β-CD) by up to a factor of 8, but it is not totally inhibited, as was supposed earlier. The effects of this modified cyclodextrin on the two series of esters are very similar, but they vary significantly with the acyl chain lengths. For example, second-order rate constants (k2) for reaction of the esters with diMe-β-CD show little change from C2 to C5, a steep increase from C5 to C7, and then a levelling off. Overall, the behaviours of γ-CD and diMe-β-CD differ significantly from those found earlier for ester cleavage by α-CD, β-CD and ‘hydroxypropyl-β-CD’. These differences are discussed in terms of the relative importance of transition-stage and initial-state binding, and in relation to the structural variations among the five cyclodextrins.

Retardation of acetal hydrolysis by cyclodextrins and its use in probing cyclodextrin–guest binding

Hydrolysis of benzaldehyde dimethyl acetal 1 in aqueous acid is slowed down greatly by cyclodextrins (CDs): α-CD, β-CD, hp-β-CD (hydroxypropyl-β-cyclodextrin) and γ-CD. The variations of the observed first-order rate constants (kobs) with [CD] exhibit saturation behaviour, consistent with 1∶1 binding between 1 and the CDs. In the case of β-CD and hp-β-CD, the binding is relatively strong and the CD-bound acetal is unreactive. In contrast, binding of the acetal by α-CD and γ-CD is much weaker, but only with α-CD does the CD-bound form show significant reactivity. The four CD-mediated reactions have been evaluated as probe reactions for determining dissociation constants of {CD– ‘guest’} complexes. In this approach, added guests attenuate the retarding effect of CD–substrate binding and cause an increase in the rate of acetal hydrolysis. The method works well for aliphatic alcohols and ketones binding to β-CD and hp-β-CD, but it is less successful with α-CD because of the shallow dependence of kobs on [α-CD] in the probe reaction. With γ-CD, the approach is not applicable at all, because added guests cause a further reduction in the rate of acetal hydrolysis, not an increase. Various implications of these findings are discussed.

CATALYSIS OF THE ENOLIZATION OF INDAN-2-ONE BY CYCLODEXTRINS IN AQUEOUS SOLUTION

In basic aqueous solution, enolate formation from indan-2-one (2, pKa= 12.2) exhibits saturation kinetics when cyclodextrins (CDs) are added, consistent with the formation of 1 : 1 complexes between 2 and the CDs. With α-CD, β-CD, γ-CD, ‘hydroxyethyl-β-CD’ and ‘hydroxypropyl-β-CD’, the reaction is accelerated up to 22-fold, but ‘dimethyl-β-CD’ slows it down by ca. 46%. All of the CDs (pKa= 12.2) are more reactive towards 2 than is trifluoroethanol (pKa= 12.4). Kinetic parameters for the CD-catalysed deprotonation are discussed in terms of the differences between transition-state binding and initial-state binding, and of the structures of the various CDs. It is concluded that anions of the CDs act as general bases towards 2, facilitated by partial inclusion of the transition state in the CD cavity, the extent of which depends on the CD. Enolate formation catalysed by β-CD is slowed by simple alcohols (propan-1-ol to heptan-1-ol), but it is not really inhibited by them, even though they bind to β-CD. Apparently, deprotonation of 2 by an anion of β-CD can still take place with an alcohol in the CD cavity, albeit more slowly.