Why are bimolecular mechanisms favored over unimolecular ones?

[David_W]

In studying Sn1, Sn2, E1, and E2 reactions; my textbook states that the bimolecular mechanisms are generally favored over the unimolecular ones. However, it doesn't elaborate much on why. Anyone care to shed some light on this?

[juanrga]

In studying Sn1, Sn2, E1, and E2 reactions; my textbook states that the bimolecular mechanisms are generally favored over the unimolecular ones. However, it doesn't elaborate much on why. Anyone care to shed some light on this?

 

A molecule alone cannot do anything. It is only when it interact with some other molecular entity, nuclei, or particle that it can change.

 

All molecular mechanisms are bimolecular. Unimolecular reactions are really bimolecular! For instance, the transformation of cyclo-propane into propene is approx. modelled as unimolecular process

 

A --> B

 

but the real mechanism (Lindemann-Hinshelwood mechanisms) is

 

A + A --> A + A* --> A + B

 

which involves two molecules.

[mississippichem]

In studying Sn1, Sn2, E1, and E2 reactions; my textbook states that the bimolecular mechanisms are generally favored over the unimolecular ones. However, it doesn't elaborate much on why. Anyone care to shed some light on this?

 

Well that can depend a lot on conditions, namely solvent conditions. The two unimolecular processes you mentioned go through a carbocation intermediate. A polar solvent helps to promote the solvation of that charged intermediate and therefore promotes the unimolecular process.

 

Now kinetically speaking, the bimolecular processes do tend to be faster as they only have one activation barrier and no intermediate with an observable half-life. If we could allow an S_N2 and an S_N1 process to compete for the same substrate and same leaving group going to the same product (in a non-polar solvent), I imagine that the S_N2 process would dominate as their is no solvent stabilization for a charged intermediate and the activation barrier for the formation of the LG-C-Nu transition state complex should in general be less than the barrier to cleave a charged leaving group away from a charged substrate in a non-polar solvent.

 

In a polar solvent it's harder to make a general statement as dielectric constants, polarizability of solvent molecules and other considerations quickly make any qualitative thoughts reduce to a battle of effects (of unknown magnitude).

 

So in short, S_N2 and E2 tend to be kinetically favored in non-polar solvents. S_N1 and E1 are favored are thermodynamically favored in polar solvents.

A molecule alone cannot do anything. It is only when it interact with some other molecular entity, nuclei, or particle that it can change.

 

All molecular mechanisms are bimolecular. Unimolecular reactions are really bimolecular! For instance, the transformation of cyclo-propane into propene is approx. modelled as unimolecular process

 

Okay fair enough but think about rate determining steps. The rate equation for a process may have just one concentration term no?

 

To be a bit more picky what about the relaxation and destruction of an exciplex or excimer? Some of these processes can be even zero order IIRC as they are only rate limited by the stability of the excited electronic state.

[juanrga]

Okay fair enough but think about rate determining steps. The rate equation for a process may have just one concentration term no?

 

To be a bit more picky what about the relaxation and destruction of an exciplex or excimer? Some of these processes can be even zero order IIRC as they are only rate limited by the stability of the excited electronic state.

 

Although the fundamental processes on nature are bi-'molecular' (so far as I know), you can get kinetics of order one, two, three, depending on the mechanism and experimental conditions.

 

I wrote molecular between quotes, because it also applies to unstable particles and radioactive nuclei. What happens is that the real de-excitation or relaxation process can be almost always linearized, as approximation, and then you get a master or kinetic equation of order one.

 

The usual kinetic equation for decay of unstable A* --> P

 

dP/dt = k[A*]

 

is derived, as approximation, to a underlying 'master' equation for a fundamental process (A* + E --> P + E') after assuming E'=E and then taking E to be in a thermal state. The Redfield equation, used in NMR, is derived in a similar way from a more fundamental bi-'molecular' expression.

Edited April 9, 2012 by juanrga

[Suxamethonium]

I understand your point, but I don't think the word "unimolecular" is explicitly meant to mean one molecule in a reaction sense, just that only one reactant is rate limiting. Like wise, bimolecular implies 2 reactants are rate limiting.

[juanrga]

I understand your point, but I don't think the word "unimolecular" is explicitly meant to mean one molecule in a reaction sense, just that only one reactant is rate limiting. Like wise, bimolecular implies 2 reactants are rate limiting.

 

Yes, I suppose that different people can be using the term in different ways.

 

I assumed that was being used in the IUPAC sense, which is (their official definition of molecularity):

The number of reactant molecular entities that are involved in the 'microscopic chemical event' constituting an elementary reaction. (For reactions in solution this number is always taken to exclude molecular entities that form part of the medium and which are involved solely by virtue of their solvation of solutes.) A reaction with a molecularity of one is called 'unimolecular', one with a molecularity of two 'bimolecular' and of three 'termolecular'.

 

It is this belief on the real existence of unimolecular elementary reactions, which I cast into doubt by technical reasons.

[Suxamethonium]

Ah, yes I see. Fair enough then.