[english]

Then if I take the example of 2-chloro-3-methylbutane, it's not a symmetric molecule. Hence despite it has 2 chiral centers, it's still chiral. Is it right? Because the mirror image cannot superimpose on the molecule itself?

Right.

[movies]

No, then you would only have one chiral center!  The 3-position would have 2 methyl groups on it.

[english]

Right.  I didn't draw that.  My mind's eye sucked that time.

 

[lavoisier]

I want to ask if a molecule has 2 chiral centers, like 1,2-dichloro-1-methylpropane, how I determine whether it is chiral by considering its three-dimensional structure?

The fact that a molecule is chiral or not doesn't depend on the presence of stereocenters, because:
1. there are lots of chiral molecules that have no stereocenters at all;
2. there are lots of achiral molecules that have stereocenters.

(By the way, this shows how incorrect is the use of the term 'chiral center': one calls 'chiral center' a point of an achiral molecule).

The only certain rule to follow (as given by k.V.) is to check if the molecule is superimposable to its mirror image. If it is, then it is achiral. If it is not, then it is chiral. More technically, one could say that molecules are achiral when they have at least one symmetry element of the second order (a rotoreflexion axis, i.e. a plane, a center... of symmetry).

I disagree with the use of the CIP rules in this case, because they'll cause confusion when you have no stereocenters or when you consider meso compounds.

I hope I didn't scramble your certainties...

[movies]

I disagree with the use of the CIP rules in this case, because they'll cause confusion when you have no stereocenters or when you consider meso compounds.

CIP rules still apply to achiral molecules.  There shouldn't be a problem.  It's the best way to define precisely which stereoisomer you are dealing with.

[lavoisier]

I disagree with the use of the CIP rules in this case, because they'll cause confusion when you have no stereocenters or when you consider meso compounds.

CIP rules still apply to achiral molecules.  There shouldn't be a problem.  It's the best way to define precisely which stereoisomer you are dealing with.

I didn't write that CIP rules don't apply to achiral molecules, and it's true that absolute configuration of stereocenters is precisely defined by the CIP descriptors. This is just stating the obvious.
I wrote that I disagree with their use in this case, i.e. for determining if a molecule is chiral or not.

Not only because the symmetry considerations are the primary reference for chirality, but also because there are at least two examples where CIP rules seem to complicate the problem.

One is atropoisomers, like many binaphthyl systems. There you have stereoisomers without stereocenters. Stereogenic axes were defined to get around the issue, but I think it's still much simpler to use the old mirror image method. Besides, with the very good freeware applications (like ACD ChemSketch), it's easy for everyone to visualise them.

The other is meso compounds. Take for instance meso-tartaric acid, which is (2R,3S)-2,3-dihydroxysuccinic acid. Someone inexperienced could erroneously think that (2S,3R)-2,3-dihydroxysuccinic acid does exist and is the enantiomer of the first! And in fact, unless one is a genius, how does he know that the two molecules are actually the same without drawing them?

But still, this just my view, someone may like CIP rules so much that all these problems don't bother him.

[movies]

I didn't write that CIP rules don't apply to achiral molecules, and it's true that absolute configuration of stereocenters is precisely defined by the CIP descriptors. This is just stating the obvious.
I wrote that I disagree with their use in this case, i.e. for determining if a molecule is chiral or not.

Ah, I see.  Yes, I agree that it is not the way to determine whether or not a molecule is chiral.  I thought you meant that it wasn't applicable at all to meso compounds.

[deutdeut]

But I have an example: Consider a compound consists of a C=C double bond. There are four groups of atoms attchong the carbons: On the top left, there's -COOH group, top right, -CH2CH3group, bottom left, a H atom and bottom right, -CH2CH3 group. This molecule should not have any element of symmetry and yet I can still superimpose the mirror image with the molecule and still it is achiral!

[movies]

That molecule has a mirror plane that contains the C=C bond and the carbon atoms of your substituents.

[lavoisier]

But I have an example: Consider a compound consists of a C=C double bond. There are four groups of atoms attchong the carbons: On the top left, there's -COOH group, top right, -CH2CH3group, bottom left, a H atom and bottom right, -CH2CH3 group. This molecule should not have any element of symmetry and yet I can still superimpose the mirror image with the molecule and still it is achiral!

Why do you say this molecule 'should not have any element of symmetry' ?
If you find it difficult to visualise it just on paper, try ACD ChemSketch: it's freeware and it gives you a very nice 3D image of your molecules. You can even draw enantiomers, rotate them in space and verify that they don't overlap.

If instead you're talking about asymmetry in conformers of your molecule, then it's a different problem, and normally in stereochemistry you consider an averaged conformer which has the highest possible symmetry, without being completely unreal (e.g., you don't draw planar cyclohexanes...).