[Rhiozan]

Dear community. For some years i had an idea of creating a software that would be capable of modeling of the behavior of the substances. Basically the idea is to create a software that would tell us the outcome of the chemical reaction between the specified elements or compounds. Surfing the web i couldn't really find anything similar to what i have been creating.

So far for me it was a matter of a hobby, I have been working on this for more than a year and this little system i have created was a subject of diploma work for my degree.

Since I'm no chemist (but learned uni-grade chemistry on my own) the main question for me now is if such engine is needed (at least for the purposes of education). Also i would like to invite the members of this community to visit my future webpage and leave a feedback about this development.

[Borek]

Strange things you are writing...

Such an engine is a sure Nobel

Predicting products is an insanely complicated task in many cases. I would say that's the problem that every chemist in at least the last three hundred years has dealt with, yet as far as I know nobody succeded.

Yet you wrote

So far for me it was a matter of a hobby, I have been working on this for more than a year and this little system i have created was a subject of diploma work for my degree.

Since I'm no chemist

And that leaves me stunned. I don't believe you were able to prepare a useful system that really works.

You may want to read this page:

http://www.chembuddy.com/?left=EBAS&right=chemical-reactions

to understand my doubts.

Please elaborate on what your system is capable of and what it is not capable of, and we will start from there.

[Rhiozan]

Thank you for the nice post, Borek. I expected such reaction  and i will explain myself little bit more.

As it is said in the link you provided, the conventional database is not a solution. When i thought of working in this field i considered and refused to make a database of reactions for the same reasons. It is extremely hard to predict the outcome of the reaction. But the point is that in the world of chemistry everything goes according to specific laws. Lots of them are known to us. Certainly there is almost always a place for probabilities and chances as well, but even this can be modelled.

My idea is to go much deeper - to model the very physics of the process. How i see it? I have created a tiny experimental database with few elements and their properties. The most important part in this is electron shell structure. I have created with PHP the most simple engine that analyzes the electron structure of reacting elements, combines them accordingly, thus emulating the covalent bonds. At this moment the engine combines any elements, regardless of their reactivity in real life, but thats because I'm myself looking for the answers and thats why joined this forum.

[Borek]

But the point is that in the world of chemistry everything goes according to specific laws. Lots of them are known to us. Certainly there is almost always a place for probabilities and chances as well, but even this can be modelled.

So far you are right - that's the general picture. I would even say that we know all laws that have to be taken into account (personal opinion, not necesarilly everyone has to agree with me on this one).

I have created a tiny experimental database with few elements and their properties. The most important part in this is electron shell structure. I have created with PHP the most simple engine that analyzes the electron structure of reacting elements, combines them accordingly, thus emulating the covalent bonds.

This is waaaaaay too simplified. I understand that's just a starting point, but my bet is that it is too simplified even for a starting pont. It even doesn't touch the important part of the whole picture.

On the very general level, chemical reactions proceed because the energy of the products is smaller than the energy of ractants. As long as you ignore energies, you won't get far. But to calculate energy you have to work out the Schroedinger equation - and that's a thing we can do only for a hydrogen atom. We don't know the exact solution even for helium. The more electrons, the more nuclei, the more problems - and all we can do is to find approximate solutions. Which sometimes help explain what is going on, sometimes not.

[Rhiozan]

Well, i had to start from something . I confess, at this point i model not the behavior of atom, but the way of thinking of a chemist who is trying to predict what is going to be as a result of the reaction.
But yes, energy and entropy are the main pushing forces of the reactions. This is the part im stuck upon.  Im trying to understand the ways of application of Schrodinger's equation in the future.

[Cheemistree8889]

Didn't Corey attempt to make a program like this for organic reactions but couldn't get the support of the chemical community?  Maybe I'm wrong but I thought I hear about it...

[Rhiozan]

Never heard of anyone trying this. The task is incredibly complicated indeed and is not only for one person. There should be a big team working on that. Too bad that community doesn't really encourage such attempts.