Polar aprotic solvents have a less hindered electronegative atom pointing out, namely oxygen, fluorine, and nitrogen, all of which have lone pairs besides having a partial negative charge. Now, all the molecules inside a reaction mixture are moving at great speeds. Not all nucleophiles will attack the carbon at 180° to the leaving group since this provides the best overlap between the nucleophile's lone pair and the C–X σ* antibonding orbital. Other nucleophiles running in random directions can maybe bounce off the solvent molecules’ oxygens’ exposed lone pair right into the antibonding orbital, thereby increasing the number of molecules that successfully overlap in some unit time. The nucleophiles are also negatively charged which increases the repulsion while 'bouncing off'. This is illustrated in the picture attached. I am aware of the present explanation which suggests that oxygen or fluorine can solvate the corresponding cations of the nucleophile and it makes sense. This is just another reason that I had thought of(there can be more than one favorable reasons for phenomena?). The number of solvent molecules will obviously outnumber the substrate and nucleophiles, the ones which are roaming around idly can maybe aid this acceleration ( vis-a-vis the repulsion hypothesis) and increase the rate doubly (not doubly but slightly more than the previously anticipated increased rate). enter image description here There is a paper that I am currently working on regarding this. I have been made to realize that I'm going to need a molecular dynamics software to check if this phenomenon is statistically more or less likely. However, I'm currently in high school, hence I do not know anything about the same. Can someone please recommend a software and guide me through the basics? Is this something that I can independently work out, or will I need some assistance from a university( or professors)? How much time and work will it take to simulate?

  • 1
    $\begingroup$ I'm voting to close this question as off-topic because you're asking us to write at least a Master's thesis. $\endgroup$ – Todd Minehardt Apr 7 '20 at 1:25
  • $\begingroup$ @ToddMinehardt what no? I'm just asking you to recommend me a software and if this is something I'll be able to do myself. this can be as short as blah blah ( software name) and yes you can or no you cannot $\endgroup$ – prarabdh shivhare Apr 7 '20 at 1:48
  • 1
    $\begingroup$ @ToddMinehardt what does closing do? I think my answer is still helpful although I agree the question is way to difficlt $\endgroup$ – Cody Aldaz Apr 7 '20 at 2:30
  • 1
    $\begingroup$ @CodyAldaz closing means that in its current form, the question is not a suitable one to be answered on SE. For example, if a question is too broad i.e. asks too much at once, then this is simply not the best place to do it (better to sit down and talk to somebody about it). This is interesting, but almost certainly too broad. It is basically planning a research project, plus asking for a guide on software, which takes tens or hundreds of pages depending on its complexity. I think you too would know that your answer, while definitely helpful, only skims the surface of what could be done. :) $\endgroup$ – orthocresol Apr 7 '20 at 9:46
  • $\begingroup$ @ToddMinehardt and orthocresol, the user says "I'm currently in high school, hence I do not know anything about the same." Therefore, do we have to close the question, or is there something we can do to save its life? $\endgroup$ – user1271772 Apr 17 '20 at 1:55

Your hypothesis concerning a "bouncing" contribution to the reaction kinetics would be incredibly difficult to observe or quantify even with the best supercomputers.

The likelihood of a molecule to react in a manner like the SN2 reaction is actually an incredibly rare event at the molecular time scale (femtoseconds 10^-15 s). This may sound contradictory since these reactions are commonplace and occur readily under appropriate conditions (e.g. a few hours heating). However, the two timescales are separated by over 15 orders of magnitude. Many reactions which do approach the transition state can also revert back to the reactant, or in other words abort. So you'd have to run the incredibly long simulations thousands of times in order to see only a small number of reactions. This is the meaning of the pre-factor in the Arrhenius rate equation

$k = Aexp(\frac{-\Delta G^\ddagger}{k_B T})$

Nevertheless, some of your intuition is correct. A molecule would indeed require momentum along the transition vector in order to overcome the highest energy point (the transition state) and pass over to the product. Otherwise it would return to reactant. Some of this velocity may naturally come from bouncing off a nearby solvent molecule.

Extra Information:

The topic concerning the relationship of transition states and molecular dynamics is actually very complex and a major challenge in the field of chemistry. The challenge essentially boils down 1) the size of the system and 2) the time length of the simulation.

Fortunately for most practical cases we can simply ignore dynamical effects and study isolated because it is known that solvent interactions are generally weak and reactions closely follow the minimum energy paths.Therefore, the free energy at the transition state ($G^{\ddagger}$) can be approximated in solution using the Harmonic approximation and implicit solvent models like PCM. If you'd like to see how some of these calculations are done, please check out my GitHub on the growing string method (ref 1). These are much simpler to model and actually give a lot of information. Usually the transmission coefficient is assumed to be 1. or 0.5.

If you're interested in some more advanced considerations of the relationship between dynamics and reaction barriers I'd recommend the following references: ref 2,3


  1. "Single-ended transition state finding with the growing string method" Journal of Computational Chemistry 2015, 36, 601–611
  2. "QM/MM Protocol for Direct Molecular Dynamics of Chemical Reactions in Solution: The Water-Accelerated Diels–Alder Reaction" J. Chem. Theory Comput. 2015, 11, 12, 5606-5612
  3. https://cims.nyu.edu/~eve2/string.htm

Not the answer you're looking for? Browse other questions tagged or ask your own question.