[Lourdes]

Hey there!
I just have an overall question concerning constitutional isomers. I have the biggest problems determining whether two compounds are structural isomers or not. So here is how I would do it: Let's say I look at a certain C-atom in one compound. And then I determine to which atoms this specific carbon is attached to. So my next step would be to see if there is ANY carbon atom in my second structure that is attached to the same atoms. If I don't find any, then these two compounds are constitutional isomers for sure. Am I right with my assumptions? Because otherwise I have no idea how to tackle this problem...

Thank you so much for helping me out here!

[Babcock_Hall]

I take as a given that you have already determined that two compounds are isomers.  If so, then your approach sounds reasonable, assuming that I am understanding it correctly.  It sounds as if you are looking at the connectivity between atoms.  If the connectivity is the same, what would you conclude?

[Enthalpy]

Shall a human determine that? It works generally without a method.
Or shall software give the answer? Then the problem gets seriously difficult. It resembles graph matching, which is known to be time-consuming.

"ANY carbon atom in my second structure that is attached to the same atoms": I guess this doesn't suffice. Checking if the atoms bonded with C are of same nature doesn't suffice. They must have the same "identity", I mean, have bonds to atoms that themselves have the same "identity", to describe the same molecule rather than an isomer.

For instance: and
Every carbon (and every atom) of left compound has a homologue in the right compound with neighbours of the same kind, that is:
carbon #2 left corresponds to carbon #3 right and #3 left corresponds to #2 right,
but the compounds are isomers, not identical. What distinguishes #2 left from #3 right is that it bonds with a carbon bearing an OH.

Other example: and
here a software can't get help from OH and Cl, and the first neighbours of -CH(CH₃)- are all -CH₂-, on the left as on the right molecule, so comparing the first neighbours doesn't suffice - nor neighbours up to some fixed distance, with a more twisted example.

In maths and software, the problem is made even more difficult by taking "atoms" of identical nature
https://en.wikipedia.org/wiki/Matching_(graph_theory)
but molecules can be huge, and then matching is a hard task. Naive methods have a complexity like N!, and complexity like N²Log(N) demands non-trivial algorithms.

[Lourdes]

Thank you guys for your replies!
Babcock_Hall, yes I looked at the connectivity. I think that if they have the same connectivity then we are talking about stereoisomers, am I right?
Enthalpy, thank you for your examples! But your second example is not very clear to me. All Carbons have the same connectivity, so these two compounds cannot be structural isomers, are they identical?

[Babcock_Hall]

If the connectivity is the same but the three-dimensional structure is different, then the two compounds are stereoisomers.

[Lourdes]

OK, got it! Thanks

[Enthalpy]

The depicted 1,3- and 1,4-dimethylcyclohexane are not identical but isomers. This shows that checking the immediate neighbours of all atoms doesn't suffice to distinguish isomers from identical compounds.

[Lourdes]

Enthalpy, how is your approach on this type of exercise? I thought checking the connectivity would be enough, can you help me figure this out?

[kriggy]

You start with molecular formula:
 if its same (then they are isomers) then you go to connectivity between atoms -if its same they are stereoisomers, if different then constitutional isomers.
If connectivity is same then you check the stereochemistry - if all stereochemistry is same then its same compound, if all is oposite then they are enantiomers. If the stereochemisry differs only on some stereocenters then they are diastereomers.

[Lourdes]

if its same (then they are isomers) then you go to connectivity between atoms -if its same they are stereoisomers, if different then constitutional isomers.

That is actually my problem. For me different connectivity means a carbon atom connected to different "neighbors". When I find a carbon in my other structure that is not connected to these "neighbors" then I have a case of constitutional isomers. Like in my first example down below.
For me it is crystal clear, that these are structural isomers. Because when you look at my highlighted carbon, you can see that it is bonded to a CH3 and this ring. But in the the other structure I can not find a single carbon that is bonded to CH3 AND a ring. So therefore they differ structurally.
But with my second example I am completely lost. I have gone through every type of isomer, and these two molecules are not identical either.
What am I doing wrong?

[kriggy]

The connectivity in those two molecules is different. One has SH group at position 2 and the other at position 3. You need to go further down the chain not only the closest atoms. Sometimes it gets quite tricky if the structure is complex.

[Lourdes]

How did you number the carbon atoms? And just to make sure, we are talking about my two separate pictures? (structure nr. 2 & 3)

[wildfyr]

Hes numbering them on the ring (1-5). He put #1 as is where the carbon chain branches off (thought I would nitpick and say the thiol should be #1). Your picture 2 and 3 are whats known as ortho and meta substituted benzene rings. Also known as 1,2 substituted (ortho) and 1,3 substituted (meta). Para is 1,4. When theres 3 substituents things we tend to drop those terms.

They are structural isomers. You made it hard on yourself by not rotating them so that at least one distinctive feature is facing the same way. Its just differing benzene ring substitution.

[Lourdes]

Thank you very much!

[Lourdes]

I am not going to start a new thread so I am going to continue here. I would be very grateful if someone could go through my exercises, so that I can make sure that I am doing this right.

-) I came to the conclusion that my first two examples are diastereomers (cis/trans).
-) I think that the next two are identical.
-) My last two are conformational isomers. By the definition of our professor conformational isomers are ones that you can get by rotating around C-C bonds. What I did here was turn the first structure (conformation1) so that the two CH3s branching off would be on the left side. And then I turned the ring so that the NH2 group and the OH group would switch their position (so NH2 on top now and OH at he bottom).
Can you guys tell me if my way of thinking is reasonable?
Thank you!