pKa training vs. change in conjugation state

Hi there,

With your help, I've been able to train pKa calculations on a number of systems with literature data and have successfully improved the predictions substantially.  Unfortunately when I recently tried to add pKa data on two acridine-like systems for which the acidic nitrogen changes its conjugation state upon loss of the proton, and cxcalc failed to generate any parameters.  I've attached the file used for training in case I've made any mistakes in the format, but I'm wondering if the pKa calculation protocol is just not capable of recognizing the effect that a changed conjugation state can have in lowering a pKa?

Thanks in advance!

                                                                                        - Gerry

 

P.S.  I'm also attaching a chemdraw file showing the two acid/base reactions that I'm attempting to account for.

Hi,

cxcalc failed to generate any parameters.  I've 

attached the file used for training in case I've made any mistakes in 

the format, but I'm wondering if the pKa calculation protocol is just 

not capable of recognizing the effect that a changed conjugation state 

can have in lowering a pKa?

 

Parameters are not generated because the number of the strucutres is not enough for a suitable regression.

May be the next changes will solve your problem.

Instead of  the 'protonated' and  'exocyclyc resonant' forms I propose to use the attached resonant forms.

 

Jozsi