Doubt cast on X-ray structure of trapped reactive species

Can 1,3-dimethylcyclobutadiene and carbon dioxide co-exist inside a supramolecular cavity?

Immobilisation inside a cavity can be a very effective strategy for stabilising reactive species. In fact, earlier this year, a team of French scientists claimed in Science1 to have used this technique to elucidate the solid state crystal structure of 1,3-dimethylcyclobutadiene trapped along with carbon dioxide inside an optimised host.

But, says Henry Rzepa from Imperial College London, UK, this claim should be treated with caution as his calculations suggest that the Science report is incorrect.

Based on his quantum chemical modelling studies, Rzepa proposes that reported crystal structure in fact is not that of 1,3-dimethylcyclobutadiene and carbon dioxide, but more probably that of the precursor used to attempt to generate the pair.

Graphical abstract: Can 1,3-dimethylcyclobutadiene and carbon dioxide co-exist inside a supramolecular cavity?

Find out more about this controversial issue in Rzepa’s ChemComm communication (free to access until 25th January 2011) and let us know what you think by leaving your comments below.

For further discussion, see Crystallographic Confusion in Chemistry Views magazine and Henry Rzepa’s blog.

1. Y.-M. Legrand, A. van der Lee, M. Barboiu, Science 2010, 329, 299-302

Digg This
Reddit This
Stumble Now!
Share on Facebook
Bookmark this on Delicious
Share on LinkedIn
Bookmark this on Technorati
Post on Twitter
Google Buzz (aka. Google Reader)

10 Responses to “Doubt cast on X-ray structure of trapped reactive species”

  1. Mihail Barboiu says:

    Irradiation of crystals containing the alfa-pyrone precursors enabled us to observe two distinct geometries of 1,3-dimethyl-Cyclobutadiene-CBD: the square planar Me2CBDS /CO2 complex (62.6%) and the rectangular-bent Me2CBDR molecule (37.4%) formed by CO2 elimination in distinct calixarene cavities, stabilized under confinement by the G4C host matrix in different orientations.

    The main controversy of this discussion is related to the short distance observed between CO2 and CBD. We agree that these distances are at the limit of the covalent bonds and we proposed an alternative model: the formation of the Me2CBDS+-CO2- O-zwitterion with a bent geometry for the CO2- moiety is more in line with the observed distances between Me2CBDS and CO2 while the square geometry in the CBD ring is more in line with the formation of the Me2CBDS/CO2 complex. These two structures Me2CBDS+-CO2- and Me2CBDS/CO2 are probably superposed in crystal.

    During late seventies, the isolation of cyclobutadiene generates similar debates on the CO2/CBD interactions in the argon matrix. They are based on the matrix splitting effects on the vibrational spectroscopic bands. Then, it was clearly agreed that this experimental evidence results “from the repulsion between CBD and CO2 rather an attraction and the strong CBD/CO2 interactions are favored by the matrix effects.”

    In this context, our experimental evidence and all these previous contributions are in contradiction with theoretical calculations presented in this communication claiming a possible recombination between CO2 and CBD!

    In the present case, such matrix effects cannot be defined by Rzepa’s models, since the reactional environment in the crystal structure is completely different!

    The reported crystal structures showed that the guanidinium cations and the sulfonate calixarene anions generate a hydrophobic/hydrophilic cavity around the encapsulated molecules. Simple inspection of Rzepa’s simplified models will reveal theoretical architectures showing different relative positions of guanidinium cations or missing neighboring calixarene anions around the reactional molecules.

    Our opinion is that such simplified theoretical models, not reliable to the real crystal structure are omitting some important geometric constraints of the confined space and moreover complex repulsive electrostatic or attractive H-bond interactions between the confined molecules and the matrix.

    Not considering a real crystal structure can certainly create “Crystallographic confusion” ….
    But do decide for yourselves….

    (Report comment)

  2. Henry Rzepa says:

    Mihail Barboiu is suggesting that the species with significant occupancy of the calixarene is in fact not a 1,3-dimethylcyclobutadiene, but a entirely new molecule, Me2CBDS+-CO2-. This would be the ionic intermediate on the stepwise Diels-Alder pathway for recombination of isolated 1,3-dimethylbutadiene and carbon dioxide. This has two issues:

    Firstly, for virtually any reasonable computational model one choses, it will be higher in free energy than either the isolated reactants, or the final Dewar-lactone product (of the Diels Alder). It could only have any chance of being detected if it were really metastable, being constrained by a significant free energy barrier on either side.

    But, secondly, calculations at almost any level predict such a species will instead have a tiny free energy barrier constraining it to such a minimum. At 175K, such a free energy barrier would have to be > ~14 kcal/mol to give it the lifetime it would need for X-ray data to be collected. Such a large barrier is however quite incompatible with an ionic species of the suggested geometry.

    It is also suggested by Barboiu that there is also significant occupancy of a species with a square geometry in the CBD ring. As discussed by Paul Schleyer, a square geometry is long known to be associated with the triplet state of cyclobutadiene, which high level calculations indicate is ~10 kcal/mol higher than a (rectangular) singlet ground state (and which is not susceptible to solvation effects). This is also true of the 1,3-dimethyl derivative. Again, it seems improbable that it could have an extended lifetime of many hours inside a cavity induced purely by the expedient of spin-forbidden decay.

    I think the strong balance of probability must favour Scheschkewitz’s suggestion of partial occupancies of the two enantiomers of the chiral Dewar lactone precursor used to attempt to generate the 1,3-dimethylcyclobutadiene and carbon dioxide. This particular suggestion had not been discussed by the original authors. Alabugin and co have also made suggestions about the photochemical conditions which I leave to them to discuss if they wish.

    (Report comment)

  3. Mihail Barboiu says:

    I think that a collegial discussion would solved many questions if a direct ethical strategy have been chosen from the beginning.

    Our last crystal structure published in Science showed the following composition Me2CBDS /CO2 complex (62.6%) and the rectangular-bent Me2CBDR molecule (37.4%) This was the last feasible structure before the x-ray data cannot be refined. This was the case of the solid state reaction for which we already considered the Alabugin suggestion. I would like to remind the point 2. of our comment of 19 December 2010 on Crystallographic Confusion in Chemistry Views magazine:
    “2.It is known that the Dewar-β-lactone intermediate gives rise to CO2 and CBD when irradiated with light of higher energy (λ<290) (J. Am. Chem. Soc. 95, 244-246 (1973). But also important, Chapman et al. showed that the same irradiation energy is equally unfavorable for the reaction completion, since the concentration of CBD rise to a maximum and then decrease as irradiation is continued.” This is the case of our experimental protocol we reproduced many times. Further experimental results will convince you that the system is not transparent to the used irradiation procedure !

    I really appreciated your models and paper in Chem. Commun. But considering 1 calixarene molecule and 4 guanidinium ions confining the active precursors as a reference simplified system is not reliable with our crystal structure! Our robust matrix is a closed confined system in which the precursors can be connected via multiple non-covalent interactions. Every reader can compare this crystal matrix and your models to understand that your “one open face” model is neglecting some complex interactions of the confined molecules with the stabilizing matrix. My opinion is that considering the whole “closed” system can change the energetic balances we are discussing. The calculations are probably right but the conclusions are pushed to far, distorting the objective reality.

    The particular Scheschkewitz’s suggestion of partial occupancies of the two enantiomers of the chiral Dewar lactone precursors have been previously discussed in our Response to TC. The second distorted enantiomer deliberately presented in a face position in his comment can be viewed in an other orientation in the Fig.2 of this paper. I would like to remember the readers that the alfa-lactone precursor is dissymmetrically trapped in the matrix and such dissymetric position in a very restrained confined space in which the Dewar lactone cannot rotate, can produce one enantiomer.

    At this end, I would like to clearly point out our position concerning the claims we addressed in this paper:
    We crystallographically observed the Dewar-lactone together with two cyclobutadiene molecules formed in a solid state as a mixture of products in a reasonable quantity and time stability to determine their structures.

    Other interpretations deliberately used by other colleagues are engaging only their responsibilities.

    (Report comment)

  4. It is good to see an open scientific debate on this matter which I hope to be constructive. It is encouraging that all of us agree now that that C-C distance of 1.5 Ang in the “square cyclobutadiene”, Me2CBDS, corresponds to a C-C covalent bond. Perhaps, now we can also agree that the even shorter 1.4 Ang C-C distance in the “rectangular cyclobutadiene”, Me2CBDR, corresponds to a covalent C-C bond as well? If we do, then both the experimental data and theoretical calculation do agree that CO2 and CBD are connected with a covalent bond. Based on the short C-C distance alone, the above statement that ” experimental evidence and all these previous contributions are in contradiction with theoretical calculations presented in this communication claiming a possible recombination between CO2 and CBD” should be retracted.
    The matter of remaining discussion is whether the 1.6 Ang C-O distances in the two “forms of cyclobutadiene” correspond to covalent bonds or not. May I point out that this distance is much shorter than a 1.79 Ang C-C bond reported by Dr. Barboiu in the original manuscript? C-O distance which is shorter than a typical C-O distance in a transition state for C-O bond cleavage agrees with the computational prediction of very fast C-O bond formation in the zwitter-ionic intermediate made by Prof. Rzepa. As I, Michael Shatruk and our co-authors suggested in our comment, the “Me2CBDS” structure fits well to the bicyclic lactone with the covalent C-O bond. A similar conclusion has been reached by David Scheschkewitz in a separate communication.
    Perhaps, the focus of discussion should be moved away from the quality of crystal data and concentrate instead on their interpretation. There are good reasons why the quality of crystal data is relatively low. Disorder may be caused by non-homogeneous reaction in the solid state, instability of reactive intermediates etc. These factors are due to the nature of this difficult experiment and not a fault of Dr. Barboui and his team. Recognizing this difficulty, we moved the discussion of the crystallographic work to the supplementary material in our comment and focused instead on the interpretation of the obtained data.
    However, the relatively low quality of data introduces significant uncertainty in the accuracy of reported interatomic distances and, thus, any paradigm-shifting conclusions based on such experiments (such as preparation of “free cyclobutadiene” or a “zwitter-ionic intermediate with a very short C-O distance”) should be considered with caution.

    (Report comment)

  5. As far as “photochemical conditions” are concerned, the situation is simple. Photochemical transformation of bicyclic lactone intermediate into cyclobutadiene is impossible under the conditions where the starting material does not absorb light. Literature experimental data agree very well with computational analysis by Prof. Rzepa that the bicyclic lactone intermediate does not absorb light with the wavelength of >320 nm used in the original Science paper by Dr. Barboiu and coworkers. Hence, it is unclear how the last step of cyclobutadiene synthesis would occur.

    (Report comment)

  6. The last comment by Dr. Barboiu (“It is known that the Dewar-β-lactone intermediate gives rise to CO2 and CBD when irradiated with light of higher energy (λ<290) (J. Am. Chem. Soc. 95, 244-246 (1973)") has a key word missing. "Only". I would fully agree with this statement in the following form: "It is known that the Dewar-β-lactone intermediate ONLY gives rise to CO2 and CBD when irradiated with light of higher energy (λ<290)". I do agree, of course, with Dr. Barboiu that cyclobutadiene may react further under these conditions and that its concentration may reach a maximum and then decrease upon further irradiation. This is not a trivial experiment!
    Another complication with the higher energy photons is that they should be absorbed by the host matrix as well. Where does the host absorb? It should be possible to check experimentally.

    (Report comment)

  7. Henry Rzepa says:

    Igor Alabugin’s last comment about where the host absorbs is a valuable suggestion for experimental study. Whilst it should not be considered in any way a definitive calculation, a TD-DFT calculation of one unit of the calixarene with the Dewar β-lactone inside was reported in my blog as having no significant absorption peaks above ~290nm.

    (Report comment)

  8. Henry Rzepa says:

    The previous comments discuss the nature of the C…O bond between the cyclobutadiene fragment and the carbon dioxide. It is worth pointing out that the calculated distance for these two atoms in the Dewar β-lactone itself, encapsulated inside a calixarene host is 1.52Å. This is significantly longer than most C…O single bonds, and does indeed suggest that the β-lactone itself may have an elongated C-O bond, with partial ionic character.

    (Report comment)

  9. Henry Rzepa says:

    An interesting aspect which has not hitherto been prominently remarked upon is the chirality of the host. It has (idealised) C4 symmetry depending on the position of the four protons at the base of the calixarene. The free energy barrier to enantiomerisation of a (simplified model) of this system appears to be around 14-15 kcal/mol (see here, with the transition state for the rotation of protons having some zwitterionic character. Of the possible guest species trapped in the cavity, the β-lactone is itself chiral, and so diastereomeric host-guest complexes are in theory possible, although as noted here, the barriers for their interconversion (via proton rotation) are relatively low. Quite what impact these dynamic processes might have on the X-ray structure determination is uncertain.

    (Report comment)

Leave a Reply