The effect of adding a chloride atom to a benzodiazepine

[Kdem]

Is there a particular reason why adding a chloride atom to a benzodiazepine makes a benzodiazepine more potent (Ki, binding affinity) ?
Of course, this isn't necessarily about therapeutic effectiveness but potency.

Examples:

Adding a chloride atom makes:

temazepam > lormetazepam
diazepam > diclazepam
nitrazepam > clonazepam
oxazepam > lorazepam

Is this a universal phenomenon ?

 

[adder]

It works universally most probably in that an electron-deficient ring at that position increases potency as both chloro and fluoro groups do the same thing, fluoro appears to give slightly more potent compounds. 2',6'-dichloro substitution supposedly also produces potent benzodiazepines, however, if chloro is put at C4', the potency drops, so some steric problem is probably there.

 

[aced126]

It works universally most probably in that an electron-deficient ring at that position increases potency as both chloro and fluoro groups do the same thing, fluoro appears to give slightly more potent compounds. 2',6'-dichloro substitution supposedly also produces potent benzodiazepines, however, if chloro is put at C4', the potency drops, so some steric problem is probably there.

One thing that has always confused me, is if the aromatic ring is stacking with say a phenylalanine residue, surely withdrawing electron density from the ring with reduce stacking interactions and thus affinity to the receptor.

 

[tryp2fun]

One thing that has always confused me, is if the aromatic ring is stacking with say a phenylalanine residue, surely withdrawing electron density from the ring with reduce stacking interactions and thus affinity to the receptor.

Not really. Electron-rich and electron-poor aromatic rings can interact strongly due to dipolar interactions, generally called "charge-transfer" complexes. The name comes from the common observation that these complexes show strong UV-visible absorption bands resulting from excitation of an electron from the pi-cloud of the electron rich ring to the pi* orbital of the electron deficient ring, resulting in an excited state ion pair.

 

[adder]

One thing that has always confused me, is if the aromatic ring is stacking with say a phenylalanine residue, surely withdrawing electron density from the ring with reduce stacking interactions and thus affinity to the receptor.

You mean an electron-deficient ring poorly interacting with an electron-rich ring? I suppose the opposite is true for the reason tryp2fun mentions. I'm wondering though how such interactions look like when two aromatic rings are both relatively electron-rich or electron-deficient, e.g. both having alkyl groups on them with other functional groups on the side-chain fairly distant to the ring. In substituted amphetamines it seems like it doesn't matter much potency-wise whether the aromatic ring has donating or withdrawing substituents. I've been wondering for some time now what would happen if one exchanged the 2'-chloro in clonazepam or a similar benzodiazepine derivative for a methyl group, would it drop the potency or increase it, or perhaps have negligible effect?

EDIT: I found this article through the wiki talk page. I'm not adept at physical chemistry, so I'm not sure if I understand it right. It says that the pi-pi interactions don't happen through pi orbital overlap (or at least it's not the major factor), then it goes on about how it's mostly about the dispersion forces.