How to predict chemical reactions?

[askingask]

I‘ve done chemistry in high school, yet I feel I still have no idea how it really works. In physics you can just calculate everything to make predictions. But in chemistry it feels like learning vocabulary, you have to learn each word and you can not just deduce everything from simple rules. I‘ve started looking into electron orbitals and so on, but I‘m still a bit confused.

So please: how do I predict what elements at what temperature using which catalyst can react to form certain molecules. And how much energy does the reaction consume or release.

 

[mjc123]

That's just how it is, I'm afraid. In principle, everything in chemistry is deducible from physics, but in practice to calculate it from first principles is far too complicated, except for the simplest systems. In practice what we do is observe what happens, and try to rationalise it after the fact using physical principles (e.g. charge distribution, dipole moments etc.) - giving the "laws" of chemistry. These rationalisations help us to predict what happens in cases similar to those observed. For organic chemistry, the mechanistic model (nucleophilic substitutions, additions etc.) has allowed a vast amount of empirical data to be organised and understood rationally, and enables us to a great extent (though not completely) to predict reactions and design syntheses. Our understanding of inorganic chemistry is much less well developed.

 

[Borek]

You probably already know the basics - if you see an acid and a base you know what kind of reaction to expect. So that's the first step in a good direction, now you just still need to learn fine prints.

In general, there is no other way than learning as many reaction types as possible and building an intuition to see how they combine in other cases. Yes, it takes years, and there are no shortcuts.

We do have a thorough theoretical understanding of thermodynamics and quantum mechanics behind chemistry (not that they are different from those in physics), and in theory we should be capable of doing calculations and predicting outcome of every reaction. In practice such a broad and general approach is not numerically feasible, we have to limit scope of calculations to get reasonable results in a reasonable time. That in turn means often it is easier to check what will happen experimentally, than by calculations.

In a way physics is simpler - in a sense that in many cases it is much easier to isolate important parts of the problem and predict the system behavior. But don't get fooled, this simplicity is often deceiving. Simple Ohm law can prove challenging in predicting current as real resistance of real conductors doesn't want to be a simple, constant number, it depends on multiple things (temperature being the most obvious, but not the only).

 

[askingask]

You probably already know the basics - if you see an acid and a base you know what kind of reaction to expect. So that's the first step in a good direction, now you just still need to learn fine prints.

In general, there is no other way than learning as many reaction types as possible and building an intuition to see how they combine in other cases. Yes, it takes years, and there are no shortcuts.

We do have a thorough theoretical understanding of thermodynamics and quantum mechanics behind chemistry (not that they are different from those in physics), and in theory we should be capable of doing calculations and predicting outcome of every reaction. In practice such a broad and general approach is not numerically feasible, we have to limit scope of calculations to get reasonable results in a reasonable time. That in turn means often it is easier to check what will happen experimentally, than by calculations.

In a way physics is simpler - in a sense that in many cases it is much easier to isolate important parts of the problem and predict the system behavior. But don't get fooled, this simplicity is often deceiving. Simple Ohm law can prove challenging in predicting current as real resistance of real conductors doesn't want to be a simple, constant number, it depends on multiple things (temperature being the most obvious, but not the only).

Is there a good rule of thumb for elements and simple molecules? And what about catalyst?

 

[Borek]

 

[Mayhem]

Even at research level, most chemists don't "predict" products. Any prediction is usually seeded by an analogous reaction found in literature, where reactants and products are quite similar in terms of reactivity.

(of course a generalization)

 

[Merlin3189]

I was quite impressed by the periodic table and the way elements in groups had similar behaviour and graduated reactivity. I'd been very interested in chem for several years before I met it, so had knowledge of a decent range of substances and could recognise where they came in the table. Maybe it wasn't so impressive for others?
Of course, though it starts off simply, you soon get to much larger atoms and up pop the transition metals. I never even got as far as Lanthanides and Actinides! But since the earlier periods are more reactive, that covers most of the reactions I encountered and was interested in.
Unfortunately a couple of years after the periodic table, suddenly it's all about carbon, hydrogen and maybe the odd oxygen or chlorine atom.

 

[Bystander]

View attachment 351598

https://rushim.ru/books/mechanizms/march6ed.pdf

Supposed to be "Seven Reactions;" organic, inorganic, sorcery...for the entire field. Takes some gray matter, but can be done.

 

[askingask]

https://rushim.ru/books/mechanizms/march6ed.pdf

Supposed to be "Seven Reactions;" organic, inorganic, sorcery...for the entire field. Takes some gray matter, but can be done.

Wait can you elaborate on this? Are you talking about these 7 reactions below?

• Combustion reaction.
• Decomposition reaction.
• Neutralization reaction.
• Redox Reaction.
• Precipitation or Double-Displacement Reaction.
• Synthesis reaction

 

[Bystander]

Wait can you elaborate on this? Are you talking about these 7 reactions below?

No. Please go through the link.

 

[askingask]

No. Please go through the link.

Already did that.
Thats some pretty hard sciency stuff.

Do basically quantum phyiscs is the answer?

I found some software called quantum chemistry software. Would that go under the umbrella?

 

[Mayhem]

Already did that.
Thats some pretty hard sciency stuff.

Do basically quantum phyiscs is the answer?

I found some software called quantum chemistry software. Would that go under the umbrella?

Quantum mechanics fundamentally underpins the course of chemical reactions, but investigating reactions at this scale is extremely laborious. It's also not something I see done in practice, as it's usually much faster to have slaves undergraduates test a synthesis out than run the mill with QM software. That said, QM is frequently used in chemistry to model and predict the behavior of molecules, which is useful for screening synthesis targets.

 

[askingask]

Quantum mechanics fundamentally underpins the course of chemical reactions, but investigating reactions at this scale is extremely laborious. It's also not something I see done in practice, as it's usually much faster to have slaves undergraduates test a synthesis out than run the mill with QM software. That said, QM is frequently used in chemistry to model and predict the behavior of molecules, which is useful for screening synthesis targets.

I guess that’s why quantum computers are of great interest to chemist.

 

[askingask]

as it's usually much faster to have slaves undergraduates test a synthesis out than run the mill with QM software.

If synthesis have to be tested, that means there is some generall understanding of how it‘s supposed to run. How is that done?

 

[Mayhem]

If synthesis have to be tested, that means there is some generall understanding of how it‘s supposed to run. How is that done?

In short, reactivity patterns have been recognized in functional groups, meaning that an aldehyde (or some other group) connected to a simple carbon chain will react similarly, but not identically, to an aldehyde connected to a complex molecular structure, and sometimes this principle doesn't hold at all. Texts such as the one @Bystander referenced tries to describe these patterns and compares them to real world examples.

Most modern synthetic chemistry is the result of decades of cumulative experience. Very rarely does a chemist "think up" a good starting point purely from intution. Most novel ideas still take inspiration from an existing system (natural or man made).

 

[Borek]

I guess that’s why quantum computers are of great interest to chemist.

The only thing common between quantum computers and and quantum mechanics calculations is the word "quantum" - other than that there is no practical connection between these things.

Thats some pretty hard sciency stuff.

I guess that's what we are trying to tell you, it ain't for a faint hearted.

 

[askingask]

The only thing common between quantum computers and and quantum mechanics calculations is the word "quantum" - other than that there is no practical connection between these things.

I thought the big problem with calculating bigger quantum mechanical systems, is that it becomes exponentially more complex. Hence why quantum computers use this exact property of entangled quantum objects to address this issue.