A newcomer to density functional theory calculations is exposed to jargon which sounds rather non-technical at first. For instance, 'charge density' is something that is actually a term in the formalism, whereas terms like 'first principles' seem adjectival. Please help me understand what they mean and their relevance as to why are they associated with the theory almost everywhere.
1 Answer
First Principles mean starting directly at the level of established science and not making assumptions such as any empirical models or parameter fitting.
With respect to DFT, EMF (Electromagnetic force) is a very strong force governing nucleus and electrons (referring to a single atom).
With that, atoms, molecules, macro molecules and materials are built up. Whatever happens in materials (be it physical deformation or its fundamental properties) comes from this basic interaction.
The law that governs these small particles and their interaction is Quantum Mechanics. The calculation methods (DFT) based on QM do not use fitting parameters from experiments or assume anything. They are solely dependent on basic laws of physics (not Newton's laws).
Thus they are named as First Principle (ab initio) Methods.
Just to give an idea: The only information these calculation methods need is the Atomic Number of the atoms involved.
-
$\begingroup$ Do not use any fitting parameters?.. What about When do I use the lattice parameters obtained computationally vs experimentally? $\endgroup$– RuslanOct 9, 2020 at 9:23
-
$\begingroup$ The calculation methods are based on quantum mechanics, therefore, do not use any fitting parameters from the experimental data as used in MD(Molecular Dynamics) and are based solely on the basic laws of physics. You can refer books.google.co.in/books/about/… $\endgroup$ Oct 9, 2020 at 9:30
-
1$\begingroup$ In the example system there is only one atomic number: 8, and there are 6 of these atoms. Their arrangement defines whether it's ozone or oxygen. $\endgroup$– RuslanOct 9, 2020 at 9:42
-
2$\begingroup$ I don't know, I wanted you to tell what exactly parameters are acceptable in ab initio calculations and what aren't. But rejecting all parameters is obviously wrong. $\endgroup$– RuslanOct 9, 2020 at 9:45
-
3$\begingroup$ @Ruslan I think you're confusing between (a) the input given to perform a calculation and (b) the theory/algorithm used to perform a calculation on the given input. The lattice parameters referred to in the question you linked is (a), the input for a calculation and in this answer, "no parameter fitting" refers to (b), i.e., the theory used for calculation. That is, the theory doesn't use empirical data to perform the computation on a given input. $\endgroup$ Oct 11, 2020 at 5:16