Sigma-minus parameters

Substituents which can withdraw electrons mesomerically from a benzene ring often deviate from an otherwise perfect Hammett plot. The course of this deviation is the extra resonance pathway for transmitting the substituent effect not felt in the ionisation of benzoic acid. For example the ionisation of phenols does not correlate well with a owing to interactions of the type in Eqn. 49. 

Perusal of the Hammett correlation indicates that meta substituents form a perfect linear relationship but the para ones deviate. A new parameter a may be defined which will restore linearity to the plot this is the value of sigma corresponding to the ionisation constant for the para substituents predicted from the meta regression line (Eqn. 50). 

A simple mechanistic use of the a correlation is the diagnosis that addition of hydroxide to the ester is the rate-limiting step in the alkaline hydrolysis of phenyl acetates. The rate constant koH correlates much better with a than with a  

Examples of good a correlations are for the alkaline hydrolysis of amino-sulphonate esters (MeNHSOj-OAr) [37], acyloxysilanes (EtjSiOAr) [38] and acyl benzene sulphonates (PhSOj-OAr) [39]. These results indicate that the ArO- bond is being cleaved in the transition state of the rate-limiting step. A typical example is illustrated (Fig. 10) for the alkaline hydrolysis of substituted phenyldimethylphos-phinates [40] where the a dependence is consistent with ArO-P cleavage in the transition state of the rate-limiting step (Eqn. 52). 

---

I) the hydrophobic term, 1-octanol/water partition coefficient, log KQyy, and two discrete variables, (2) the ability to accept hydrogen, Ha, and (3) the ability to donate hydrogen, Hd. Several electronic parameters were selected (4) the Hammett sigma constant for the para position (see Jaffe 1953), and its special cases, (5) (normalized), (6) (minus), and (7) (plus) (see Hansch and Leo 1979) (8) the polar or field parameter, F (Swain and Lupton 1968) and (9) the resonance parameters, R (Swain and Lupton 1968), (10) R°,... 

The first out-of-control condition is when the process generates part parameters outside of some predetermined distance from the center of the output distribution. These limits are sometimes referred to as the upper process limits (UPL) and the lower process limits (LPL). Some authors refer to these limits as the upper and lower natural process limits. The selection of these limits is arbitrary. The current custom is to select plus and minus three sigma limits. If the distribution of the process output is a normal distribution, then 99.74% of the process will typically be between the upper and lower three sigma process limits. In some cases, such as for extremely stable processes, it is more practical to use 3.5 or 4.0 sigma limits instead of 3.0 sigma limits. This minimizes the number of false out-of-control signals. 

---