Recently I posted a response to a comment to the judgment that ChemSpider should not list the logP value associated with Calcium Carbonate. Our belief is that the ACD/LogP product is appropriately predicting a logP value for carbonate.Â
That said we thought it appropriate to consult with an expert. Who better than Chris Lipinski. Chris is the author of the “Rule of Five” a simple algorithm for predicting drug compounds likely to show oral activity. At Pfizer, from 1970 to 1990, he supervised medicinal chemistry drug discovery laboratories discovering multiple gastrointestinal and diabetic clinical candidates. In this process he became interested in the design of bioisosteres and in drug physical chemical properties and quantitative structure activity relationships, especially as they related to problems of oral activity.Â The comments below are an excerpt from an email exchange with Chris: â€œCarbonate anion is distinctly different than other inorganic anions in that it is only stable at base pH. The reported first acidic pKa of around 6.1 for H2CO3 is actually a composite of two numbers; the true pKa which is around 4 and the equilibrium constant between CO2 dissolved in water and true H2CO3 which favors the dissolved CO2 by a factor of about 100. It takes a distinct period of time for the equilibrium to occur on the orders of 10’s of seconds at 37 deg C. Hence carbonic anhydrase in human physiology. The stability of H2CO3 is tremendously dependent on pH, on stabilizing cations and by anything with a transient tendency to form an anhydride or mixed ester with H2CO3, eg. orthophosphoric acid in soda pop. So a complex answer to a simple question. My position would be that since universally the complex equilibria is rolled into the pKa (and presumably log P of H2CO3) I would simply list your calculated LOG P for CaCO3.â€Â
With Chrisâ€™ comments we will be leaving the logP value for Calcium Carbonate in place. We do welcome any further feedback though. We know there are errors on ChemSpider as they are being curated now by ChemSpider users as shown here. There are > 10 million records..we do need your feedback!