Boiling points

wangasu

Hi, does anybody know how to predict the boiling points of chemicals? Is there is a generally accepted theory or model to explain Tb? Thanks.

Bystander

Short answer? No.

There are special cases, homologous series, well characterized functional group effects, and a couple others that escape me at the moment, in which known boiling points for members of a series allow reasonable estimates for remaining unmeasured members.

ab initio calculation of a boiling point or vapor pressure? Do it, and you've got your Nobel sewed up tight.

Artermis

depends on a lot of things. most questions will focus on a particular aspect; like bystander has said, like functional groups, the compacity of the molecule, the formula weight, the polarity, it depends on a lot of factors...

Nerro

In short the boiling point of a substance is set by the attraction it applies upon other particles around it.

Functional groups might cause the vanderwaals bonds to be tighter because of unevenly spread charge. Mass is important too because the amount of mass a particle has decides how much gravity it exerts on other particles and sterical effects such as entanglement (not the QM variaty but just plain old hydrocarbons wrapping around eachother ;)).

All in all many many such factors must be included in the calculation when you want to calculate its boiling point.

gravenewworld

the chemical potential of a solution also plays a role in determining the boiling point.

wangasu

Thank all of you.. I am not doing calculation work. my interest is largely in liquid dynamics. one more question, do you happen to know that, upon boiling, roughly, how broad the distribution of the clusters which escapes from liquids might be? I wonder if single molecule is the only form.

Bystander

Only for molecules taking introductory chemistry or physics. Population distributions of cluster sizes are again going to depend upon intermolecular potentials. Helium? Not a whole lot of dimers, even fewer trimers, .... Water? Lots of dimers and higher order clustering. Even numbers preferred over odd? Or vice versa? Certainly --- depends on the substance.