# Deriving Chemistry From Physics

*Most recent answer: 12/12/2012*

Q:

Hello,
I don't mean to sound rude, but if Chemistry is actual fact then why does every element or every material have to be studied/experimented with separately to get information about that element or material? If the explanations of electrons and orbits and energy bands ect is factual, then why can't we simply make predictions for uninvented materials based on the information we already gathered from known materials? I understand that it can be guessed with relative accuracy what material properties will be like and such, but why is there no certainty?

- John (age 32)

Cleveland, OH, US

- John (age 32)

Cleveland, OH, US

A:

The answer is simple- the quantum equations governing multi-electron systems are hard to solve. Various approximate techniques work well, but chemical stability can depend on small differences between rather large energies.

If the electrons didn't interact with each other, this would not be an issue. The states of nearly definite energy for singe electrons around a nucleus can be found by solving simple equations with great precision. However, each electron is influenced not only by the nucleus but also by the other electrons. The various approximation techniques (Hartree-Fock, etc. ) developed over the years do a good but not perfect job for atoms. For molecules, things get more complicated still. So there's nothing wrong with just using some approximate empirical rules for cases where solving the basic equations accurately enough isn't yet possible.

Often, predictions for new materials work very well. Drugs, for example, are now often "designed". Various magnetic and superconducting materials are tweaked knowing properties of the ingredients. Some day we may be able to look at a crystal structure and make a good estimate of the superconducting transition temperature, but for now that's just too complicated.

Mike W.

If the electrons didn't interact with each other, this would not be an issue. The states of nearly definite energy for singe electrons around a nucleus can be found by solving simple equations with great precision. However, each electron is influenced not only by the nucleus but also by the other electrons. The various approximation techniques (Hartree-Fock, etc. ) developed over the years do a good but not perfect job for atoms. For molecules, things get more complicated still. So there's nothing wrong with just using some approximate empirical rules for cases where solving the basic equations accurately enough isn't yet possible.

Often, predictions for new materials work very well. Drugs, for example, are now often "designed". Various magnetic and superconducting materials are tweaked knowing properties of the ingredients. Some day we may be able to look at a crystal structure and make a good estimate of the superconducting transition temperature, but for now that's just too complicated.

Mike W.

*(published on 12/12/2012)*