Four- and five-component molecular solids: crystal engineering strategies based on structural inequivalence

A logic driven synthetic approach is used in this first report of the isolation of stoichiometric four-component molecular solids. A possible extension to a five-component solid is also described.

Considering that TMHQ is very prone to oxidation, and the crystal of the suspected quintinary is brown-yellow, it is concluded that TMHQ gets oxidized in solution and the TMBQ that is produced enters the crystal in the disordered MRE site. TMBQ has the required topological similarities to enter the disordered MRE site and facilitates the shape-size mimicry that was anticipated for TMHQ. The two orientations of TMBQ are given below: It is noted that TMBQ is bound in the site with C-H…N hydrogen bonds rather than the expected O-H…N bonds with MRE. This is a weaker interaction and we conclude that the amount of TMBQ in the site will be much less than MRE. It cannot compete so well with the MRE but it still enters the crystal to some small extent justifying the design strategy of exploiting the lack of degeneracy of the D site in the quaternary cocrystal. There is a precedent for this kind of behavior (strong versus weak hydrogen bonding) and we have observed similar solid solutions in the system barbital-urea-acetamide (Ref. 29).It is not clear why TMHQ does not enter the crystal (it is present in large excess in the mother liquor). Perhaps, it is a steric factor that allows for the entry of the TMBQ molecule and not TMHQ. Supporting information, sup-12 this site, the expected SOF will be 1.0 for the O-atom and 0.75 for each of the four C-atoms. From the observed numbers in the refinements, it seems likely that both MRE and TMBQ are present at this disordered site in some statistical manner.
The above scheme shows the SOF of the C/O atoms assuming 100% MRE occupancy and 100% TMBQ occupancy. Neither set of numbers in the actual refinements correspond exactly to the ideal numbers for 100% occupancy for each of the components MRE and TMBQ. However, both refinements are stable and no atom becomes non-positive definite. The electron density maps below convey this: Since both models are reasonable (MRE 100% and TMBQ 100%) to some degree, we conclude that the site contains both molecules.
The final ORTEP figure is shown below:

S3.2. Confirmation of the presence of TMBQ in the disordered site
Similar refinements were carried out (100% MRE and 100% TMBQ in the central inversion site) for the data of the yellow (quaternary) crystal. In this case, it was found that the refinement with 100% MRE proceeded satisfactorily (as expected). Refinement with 100% TMBQ (which we know is patently incorrect) did not proceed satisfactorily. The refinement was unstable and several H-atoms became non-positive definite. This indicates that the data on both the yellow and IUCrJ (2016). 3, doi:10.1107/S2052252515023945 Supporting information, sup-14 brown crystals is good enough to show that some TMBQ is present in the brown crystal. The data is at a better level than where either or both MRE and TMBQ may be placed at the inversion site in either data set and refined normally. INS files for the datasets (in text format) are provided for the reviewers.
We estimate that the amount of TMBQ at the disordered site is around 5% (MRE is 95%) because there is no evidence of any C=O stretching absorption in the IR spectrum of the quintinary cocrystal.

Compounds
Ratio Result