This molecule does not exist!

[Acru]

After some messing around, I managed to make an isomer(?) of Cyanogen that Sokobond says does not exist~

Screenshot: http://steamcommunity.com/sharedfiles/filedetails/?id=335954594

[Draknek]

Well done! There are a few other levels you can solve in similarly unintended ways, although I don't have a complete list to hand. See if you can find them all!

The interesting question is: why can't that molecule exist? The rules of Sokobond make a lot of simplifications that aren't true in the real world, so what is the crucial difference?

(Don't try to find them all, there's nothing too exciting.)

[Acru]

I don't know enough about real molecular bonds to say why it can't exist, though I might guess such a structure would be somehow unstable. The wiki page mentioned other possible isomers existing but not what forms.

[mat.duf]

If I remember well high school classes I would say that double bonds are very strong and reject the other bonds. This present structure can't happen because no atoms could find a way to fit.

In the same way you can't have O3 with a triangle loop (only simple bonds which would break themselves).

However some "hardly possible structures" can exist but decay very quickly.

Personally I found a level where you can have two different results.

[Sam]

I found one of those in the upper-right blue area. There's one that has about 10 molecules, and I was just experimenting -- but after about 10 moves, I had already solved it. It was a huge, weird shape, and came up as ??? as well!

[c_perfringens]

Molecular bonds have to respect some angles, in your molecule, the ring strain would be too important, the molecule would be highly unstable

[Berahlen]

Yeah I just did that solution in "Hallway" (lower-right corner).

When you start accumulating multi-bonds, the molecule sort of flattens out and the bond angles broaden to keep it stable. Packing a four multi-bonded atoms into a planar square would...well, it would want to explode really badly.

[JsKingBoo]

Right Angle alternative solution: https://www.youtube.com/watch?v=fvLa6UywaxA&feature=youtu.be

Image for those who don't want to open Youtube: http://i.imgur.com/PTCHNNE.png

[DariusGalaxyz]

Okay I'm gonna answer the Developer's question, and maybe they'll be nice and deposit some Steam Wallet funds for this because nobody else got this correct, lol. The crucial difference between Sokobond and real-World physics is the electrons. In the real-World, the electrons is what binds the atoms together to form molecules. A real Nitrogen atom has 7 electrons, whilst Sokobond Nitrogen displays only 3 electrons, hence why you can build nonexistent molecules in Sokobond. The game would have to be much much more complex if it were to mimic the real-World, it would have to be 3-D for one thing, and even then it's complicated especially for the real molecules.

[Berahlen]

Yeah but unpaired valence electrons are usually the only really important ones, and nitrogen only has three of those. Though there are ways to force bonds onto normally-paired ones, which lets you do funky stuff with sulfur, halogens, and even noble gases.

The main reason you wouldn't see the OP's molecule in real life is that twisting bonds -- especially multi-bonds -- into unusual angles is like bending a really, really explosive spring. Square loop setups like cubane are possible, but very difficult to synthesize without losing your eyebrows/limbs. Cubane is actually unusual in that its final state is relatively not explodey.

[Prime Seeker]

Another one.
https://steamcommunity.com/sharedfiles/filedetails/?id=1454055025
Not sure why this cannot exist.

[pasa]

Originally posted by Draknek:

The interesting question is: why can't that molecule exist? The rules of Sokobond make a lot of simplifications that aren't true in the real world, so what is the crucial difference?

Molecules are in 3D and as others said certain spacial angles will form as you create bonds.

Try to build this molecule: C=C is a good start and you'll have the next attachments at 120 angle like in ethilene. You put N on each with single bond, they can't reach each other to form a circle already. If you somehow force those bonds double: -N=C=C=N- it will form a straight line you can not bend (I think).

[i.baram]

I'll take a crack at explaining why that molecule can't exist. Molecules have their atoms bonded together by molecular orbitals (MOs), which are regions of space that electrons occupy. In a simplistic sense, MOs can be thought of as combinations of overlapping atomic orbitals, which are the regions around each atom that the electron can be found. Because electrons carry the same charge and repel each other, each atom's atomic orbitals will occupy regions of space as far away from each other as possible. When atoms bond, their atomic orbitals are thought of as hybridizing into hybrid orbitals that optimize electron spacing and bond energy. This doesn't actually describe what's happening, but it's a useful tool chemists use to think about bonds and to predict structures.

When an atom, like both carbons in this molecule, only has two orbital sets it needs to worry about (in this case the orbitals bonded to the other carbon and the nitrogen, because carbon has no left over electrons after the bonding), it is said to adopt "sp" hybridization. This means that two hybrid atomic orbitals form 180 degrees apart from each other in a straight line, and two more unhybridized orbitals called p_x and p_y exist each at 90 degrees to the bond axis in 3d space. One each of the p_x or p_y orbitals is used to form the second bond of the two double bonds on the central carbon.

This means that given the hybridizations of the carbons, the molecule should exist in a linear form .N=C=C=N.

For the nitrogens to bond to each other would induce incredibly strong stress in the bond, and is functionally impossible. To the best of my knowledge, no molecule with close to that amount of strain has been isolated. If you ever attempted to make this molecule it would basically instantly decay to cyanogen

I hope this made sense, this content is basically the condensation of the important bits from 1-2 lectures worth of chemistry.

[J Pystynen]

Just built "methanediimine" (HN=C=NH) in level 'Workshop' and it's registering as also cyanamide, instead of "???" or an alternate solution. (A perfectly possible molecule, even if an unstable one.)