13
$\begingroup$

I find the experiments to be relatively easy to explain to my young cousins (who are planning to take scientific reserach as a career option). I find the theoretical works, especially the ones which require highly involved computational tools (read DFT) to be a bit difficult to explain to these 16 somethings. They have a reasonably good understanding of the structure of the atom but nothing more than that.

$\endgroup$
  • $\begingroup$ Related: physics.stackexchange.com/q/524722/140194 $\endgroup$ – Verktaj Jun 11 at 19:38
  • 1
    $\begingroup$ I have not voted to close, but I do strongly recommend that you see the post in the comment by AndrésRobles-Navarro, since that is essentially asking the same thing, so the question is sort of a duplicate. Rashid has pointed out though, that duplicate questions are not necessarily a bad thing, since we can get different answers here than there. I particularly don't encourage such duplication, but I won't vote to close either. $\endgroup$ – Nike Dattani Jun 11 at 19:51
  • 1
    $\begingroup$ I think that the questions are slightly different. From the link in the @AndrésRoblesNavarro comment we have: "If you were to explain them to a person that is not necessarily familiar with quantum mechanics, how would you do it?" whereas the question here is for high school students. $\endgroup$ – Camps Jun 11 at 23:11
10
$\begingroup$

I strongly recommend you to visit the site of the North Carolina High School Computational Chemistry Server.

They have a project for teachers and high school students to learn molecular simulations using DFT, Molecular Mechanics and Semi-Empirical methods.

| cite | improve this answer | |
$\endgroup$
  • 2
    $\begingroup$ That's a great initiative! I'll wait for other answers as well so that we can enough number of different perspectives. Thanks. $\endgroup$ – Sufyan Jun 11 at 16:45
  • $\begingroup$ This looks to me more like "Here's how to run calculations" than "This is what density functional theory is". $\endgroup$ – Susi Lehtola Jun 21 at 7:02
8
$\begingroup$

Explaining DFT even to a master's student is pretty hard... and to a high school student? Lol. In that case you're really restricted to a very ambiguous description. Unlike a master's student, they don't know (and you can't show them) what the exchange term is. Moreover, they don't (really) know what a wave function or state is.

Because of this, you're reduced to a very fuzzy and ambiguous explanation, so it would help a lot knowing what context you are working in. Mentioning DFT might even be counterproductive.

The single-particle Schrödinger isn't that hard, and can be understood in terms of the wave equation, e.g. by oscillations of a string. The problem is that if you want to also involve DFT, you need to explain why you need it. This is due to the many-particle aspect: exchange and correlation. The first arises already with a single electron configuration (in addition to the classical Coulomb electrostatics you also get a term corresponding to interchange of the particles), while the latter arises from the fact that in reality many electron configurations contribute to the wave function simultaneously, and the number of possible configurations increases rapidly with system size.

DFT constructs an effective independent particle approximation for this interacting system, and the exchange-correlation functional does the magic of mapping the systems to each other. Although we don't know what the exact functional is, hundreds of approximations have been developed for it, but the problem is that it's almost impossible to tell beforehand whether the approximate functionals are reliable for the system you want to study.

| cite | improve this answer | |
$\endgroup$
  • $\begingroup$ I've learned differential equations as a high school student. I knew no quantum mechanics until my freshman year, but it doesn't sound out of the realm of possible. $\endgroup$ – svavil Jun 12 at 15:00
  • 1
    $\begingroup$ But you don't need differential equations to learn DFT. The problem is understanding many-particle quantum mechanics. $\endgroup$ – Susi Lehtola Jun 12 at 16:11
  • $\begingroup$ my DFT is rusty, but don't you need time-independent Schrodinger equations? $\endgroup$ – svavil Jun 12 at 16:12
  • $\begingroup$ I expanded on the answer above a bit. Time-independent is enough, but that's not the problem; it's the many-particle aspect. $\endgroup$ – Susi Lehtola Jun 13 at 10:47
1
$\begingroup$

In quantum mechanics state is represented by wave function while physical observables (such as energy, position, momentum etc) as operator. In quantum chemistry we are interested in obtaining energy of molecules, which is obtained as expectation value of the Hamiltonian. DFT is based on the realization that one does not need full information of the wave function as the Hamiltonian involves just one- and two- body operation in the process of evaluating the expectation value of the Hamiltonian, most of the degree of the freedom will integrate out. Thus one-particle and two-particle density will be suffice to obtain energy of the molecule. Two particle density can be written as a product of one-particle densities and a correlation function whose form is not know exactly but can be constructed using physical insights.

| cite | improve this answer | |
$\endgroup$
  • 3
    $\begingroup$ I feel like this explanation would be confusing to some undergraduates, never mind high school students. $\endgroup$ – LUPHYS Jun 12 at 13:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.