6
$\begingroup$

The $\ce{^{31}P}$ NMR spectrum of $\ce{Me2NPF4}$ when recorded at room temperature consists of a quintet with intensities $1:4:6:4:1$. However, at $-100~\mathrm{^\circ C}$, the spectrum is as shown below:

NMR spectrum at low temperature

Propose a structure for $\ce{Me2NPF4}$ that is consistent with these observations.

The phosphorus has five groups surrounding it so it must have some sort of five-coordinate geometry. The low temperature spectrum shows a triplet of triplets with similar coupling constants, which is what you would expect from a trigonal bipyramidal structure with two axial and two equatorial fluorines. On this basis I assigned the structure as below:

Structure of Me2NPF4

However, I don't understand why all four fluorines are equivalent at higher temperatures. Does this mean that they are constantly interchanging and if so, what is the mechanism for this?

$\endgroup$
4
  • 4
    $\begingroup$ en.wikipedia.org/wiki/Berry_mechanism $\endgroup$
    – permeakra
    Jan 2, 2016 at 18:14
  • 3
    $\begingroup$ See "Suggest possible structures for the compound PF3(Ph)(NMe2) using the NMR data" for some background and follow the link to Berry pseudorotation. $\endgroup$
    – ron
    Jan 2, 2016 at 18:18
  • $\begingroup$ What's really bothering me about this NMR spectrum is the lack of $\ce{H-P}$ coupling. I made a number of phosphonate compounds of the general type $\ce{RP(O)(OCH3)2}$ for my PhD, and the methoxy groups showed $^3 J_{\ce{H-P}} \approx 10-20\ \text{Hz}$. In just about any structure I could propose for $\ce{Me2NPF4}$, there should be some $\ce{H-P}$ coupling. $\endgroup$
    – Ben Norris
    Jan 2, 2016 at 23:50
  • $\begingroup$ @BenNorris The lack of $\ce{H-P}$ coupling is due to quadrupolar relaxation by the intervening nitrogen. See my comment to Jan's answer for more detail. $\endgroup$
    – ron
    Jan 3, 2016 at 10:18

1 Answer 1

2
$\begingroup$

Obviously, in the structure you drew you have two sets of two fluorines which are non-identical: The axial ones that form a 4-electron-3-centre bond with phosphorus and the equatorial ones that are bound by the classical 2-electron-2-centre bond. And also obviously, at higher temperatures this difference vanishes so the fluorines must somehow interconvert into each other (that is the easier explanation when compared to a different structure and is thus preferred by Occam’s razor) rapidly on the NMR-timescale. You noticed this already.

The mechanism for this is called Berry pseudorotation, and it involes a square pyramidal transition state as shown in this image taken from the Wikipedia page:

Berry-pseudorotation

(The image is drawn with pentacarbonyliron but just exchange iron with phosphorus, four carbonyl groups with fluorine and the fifth with dimethylamine and voilà.)

Since this mechanism involves a significant change in the binding situation (we no longer can write two clearly 2-e-2-c and one 4-e-3-c bond(s)), we need a certain activation energy — that is usually supplied thermally. At $-100~\mathrm{^\circ C}$, there is not enough thermal energy to induce this change; at room temperature there is.

Ben Norris mentioned another important observation in a comment: There is not $^3J_\ce{PH}$ coupling visible. I suspect that this is due to the spectrum actually being $\ce{^{31}P\{^1H\}}$ — a proton-decoupled phosphorus spectrum.

$\endgroup$
3
  • 3
    $\begingroup$ The absence of $^3J_\ce{PH}$ coupling is due to the fact that the coupling path involves the nitrogen and $\ce{^14N}$ is a quadrupolar nucleus - quadrupolar relaxation effectively washes out the coupling. Jan, I suspect you understand this, but for others who might like to read a bit more about quadrupolar relaxation and when it will and will not effectively decouple nuclei, here are links to two earlier answers on the subject (ref_1, $\endgroup$
    – ron
    Jan 3, 2016 at 10:14
  • $\begingroup$ ref_2) $\endgroup$
    – ron
    Jan 3, 2016 at 10:15
  • $\begingroup$ @ron Actually, I didn’t know about that. I hope I remember to read your references soon (kinda bad time atm) and then hopefully correct my answer accordingly. Thanks for pointing out! $\endgroup$
    – Jan
    Jan 3, 2016 at 15:45

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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