A chiral compound that crystallizes as a conglomerate — the two enantiomers spontaneously sorting into separate enantiopure crystals, which enables resolution by preferential crystallization or deracemization — can later reveal a more stable racemic compound that displaces it, a 'disappearing conglomerate' that invalidates a committed crystallization-based separation. Three primary sources frame the problem: Hoquante et al. (2019) documented the phenomenon for BINOL-OBn and proposed a four-part risk-assessment procedure (vary crystallization conditions; change synthesis route and impurity profile; seed with related racemic compounds; computationally screen candidate racemic structures in common centrosymmetric space groups with Z′>1Z'>1); Chen et al. (2019) documented a protonation-triggered conglomerate-to-racemic switch in a Tröger's base cryptand (a cage-like host molecule built from V-shaped chiral amine scaffolds); D'Oria, Karamertzanis and Price (2010) showed that computed racemic-versus-homochiral lattice-energy differences for conglomerate formers are typically only a few kJ/mol, so prediction requires free energies and kinetics beyond 0 K crystal structure prediction. No work since has validated the proposed risk-assessment criteria on any independent chiral system. Later studies advanced the energy-difference methodology (Otero-de-la-Roza et al. 2016; Borchardt-Setter and Yu 2023; Pinetre et al. 2025) and showed that enantiopure crystals can be kinetically stabilized far from equilibrium, but the most recent of these explicitly characterizes prior energy-difference analyses as incidental reports on structurally unrelated molecules. Determining the thermodynamic, chemical, and kinetic conditions under which a conglomerate switches to racemic-compound crystallization, and establishing a validated way to predict and control that risk, remains open.

Background

Most syntheses of chiral molecules deliver equal amounts of the two mirror-image forms, the R and S enantiomers, as a racemate. When such a mixture crystallizes, two outcomes are possible. In a racemic compound, every crystal contains both enantiomers in a 1:1 ratio, usually packed around an inversion centre in a centrosymmetric space group; in a conglomerate, every crystal contains only one enantiomer, so a racemic melt or solution spontaneously resolves into a mechanical mixture of enantiopure crystals ('spontaneous resolution'). Conglomerate formation is the enabling condition for direct crystallization-based enantioseparation — preferential crystallization (repeatedly harvesting one enantiomer from solution) and Viedma ripening / attrition-enhanced deracemization (grinding a slurry of conglomerate crystals while the solution racemizes, until a single enantiomer remains) — yet only an estimated 5–10% of chiral compounds form conglomerates, and which ones do is essentially unpredictable.

Two structural conventions are needed. Z′Z' is the number of chemically independent molecules in the crystallographic asymmetric unit; racemic compounds with Z′>1Z'>1 pack several independent, often conformationally distinct, molecules, and such structures are harder to nucleate. Crystal structure prediction (CSP) generates plausible crystal packings and ranks them by lattice energy, the computed 0 K enthalpy of the crystal relative to isolated molecules.

The reliability question. A conglomerate-based separation is only as reliable as the conglomerate itself. Solid-state chemistry documents 'disappearing polymorphs' — crystal forms that become unreachable once a more stable competitor has nucleated — and the same threat applies to conglomerates: a compound reported and used as a conglomerate may later reveal a more stable racemic compound that displaces it. Hoquante, Sanselme, Rietveld and Coquerel (2019) documented a complete case: 2'-benzyloxy-1,1'-binaphthalene-2-ol (BINOL-OBn), twice reported as a conglomerate, in fact forms a racemic compound (space group P21/nP2_1/n, Z′=3Z'=3, containing hydrogen-bonded dimers absent from the enantiopure crystal) whose melting point lies about 20 °C above the conglomerate eutectic, making the racemate the stable phase at every temperature examined; a 1:1 mixture of enantiopure powders held in diethyl ether at −10-10 °C with cross-seeding converted entirely to the racemic compound within 48 h, and the original conglomerate could not be re-obtained. Chen et al. (2019) showed that the switch can be chemically commanded: a diamond-like cryptand (a cage-like host molecule) built from Tröger's base units (V-shaped, inherently chiral amine scaffolds) that crystallizes with chiral self-sorting as a conglomerate (space group C2221C222_1) crystallizes instead as a racemic compound (space group P1ˉP\bar{1}, both enantiomers per crystal) after mono-protonation with two equivalents of trifluoroacetic acid. The same vulnerability constrains practice: the conglomerate salt of 2-phenylglycinamide with N-acetyl tryptophan converts under Ostwald's rule of stages (metastable forms appear first) to more stable phases, closing the processing window for its deracemization (Lerdwiriyanupap et al., 2022).

Why prediction is hard. D'Oria, Karamertzanis and Price (2010) computed complete crystal energy landscapes — both observed homochiral and hypothetical racemic packings — for three known conglomerate formers and found the decisive lattice-energy differences are typically below about 4 kJ mol−14\ \mathrm{kJ\,mol^{-1}}, with racemic and homochiral packings often nearly isostructural. Such margins are comparable to vibrational-entropic and thermal effects and can change sign with temperature, so a 0 K lattice-energy ranking cannot decide which phase wins under crystallization conditions; relative nucleation and growth kinetics (metastable conglomerate first, stable racemate later) must also enter. Hoquante et al. (2019) therefore proposed a practical four-part procedure to anticipate the threat before process commitment: (i) challenge the conglomerate under as many crystallization conditions as possible (temperature cycling, ultrasound, different solvents and supersaturations); (ii) change the synthesis route, because impurity profiles alter nucleation and growth rates; (iii) seed with racemic compounds of related chiral molecules; and (iv) computationally screen candidate racemic structures in common centrosymmetric space groups (e.g. P21/cP2_1/c, P1ˉP\bar{1}, C2/cC2/c) with Z′>1Z'>1 and assess their energies against the enantiopure crystal — stressing that time matters, since the metastable conglomerate is kinetically favored. As proposed, this procedure was demonstrated on no system other than BINOL-OBn itself, and its authors note that recovering a displaced conglomerate requires finding a solvent or impurity that suppresses nucleation of the stable racemic phase — a condition currently discovered only empirically.

Problem Statement

For a chiral compound that crystallizes as a conglomerate, determine the conditions under which crystallization switches to the racemic compound — a 'disappearing conglomerate' — and establish a validated way to predict and control that risk before a crystallization-based resolution process is committed.

The open problem spans three linked aspects: (1) Conditions — identify which factors govern conglomerate stability in a chiral system: the free-energy difference between the racemic and homochiral crystal phases as a function of temperature, the chemical state (protonation or salt formation, solvent, coformer, synthesis route and impurity profile), and kinetic factors (relative nucleation and growth rates, including long-delayed appearance of the racemic phase), and show by experiment or validated computation how these factors decide whether the conglomerate persists or is displaced. (2) Prediction — criteria or a methodology for anticipating disappearing-conglomerate behavior, extending the four-part procedure proposed by Hoquante et al. (2019) (challenging the conglomerate under many crystallization conditions; varying synthesis route and impurity profile; seeding with related racemic compounds; computationally screening candidate racemic structures in common centrosymmetric space groups with Z′>1Z'>1), and validated on chiral conglomerate-forming systems independent of those from which the criteria were derived. (3) Control — conditions (for example a solvent, additive, or impurity that suppresses nucleation of the racemic compound) that preserve or recover a metastable conglomerate, or that deliberately induce or reverse the switch, demonstrated on a system where that direction of control is not already documented.

The single overarching objective is reliability of spontaneous resolution: converting the present case-by-case understanding into a predictive and controllable assessment of conglomerate vulnerability. Contributions that address only one system, one energy computation, or one kinetic stabilization trick are partial results; a general, cross-system-validated framework for predicting conglomerate displacement, and levers to control it, do not yet exist.

References

  1. Marine Hoquante, Morgane Sanselme, Ivo B. Rietveld, Gérard Coquerel. Disappearing Conglomerates, Assessment of the Threat. Crystal Growth & Design, 2019, 19(12), 7396–7401. DOI: 10.1021/acs.cgd.9b01316. https://doi.org/10.1021/acs.cgd.9b01316 (full text also available from HAL, https://hal.science/hal-02336086)
  2. Yuan Chen, Cheng Qian, Qian Zhao, Ming Cheng, Xinran Dong, Yue Zhao, Juli Jiang, Leyong Wang. Adjustable chiral self-sorting and self-discriminating behaviour between diamond-like Tröger's base-linked cryptands. Chemical Communications, 2019, 55(56), 8072–8075. DOI: 10.1039/c9cc03577j. https://doi.org/10.1039/c9cc03577j
  3. Emiliana D'Oria, Panagiotis G. Karamertzanis, Sarah L. Price. Spontaneous Resolution of Enantiomers by Crystallization: Insights from Computed Crystal Energy Landscapes. Crystal Growth & Design, 2010, 10(4), 1749–1756. DOI: 10.1021/cg9014306. https://doi.org/10.1021/cg9014306
  4. Abraham Otero-de-la-Roza, Jason E. Hein, Erin R. Johnson. Reevaluating the Stability and Prevalence of Conglomerates: Implications for Preferential Crystallization. Crystal Growth & Design, 2016, 16(10), 6055–6059. DOI: 10.1021/acs.cgd.6b01088. https://doi.org/10.1021/acs.cgd.6b01088
  5. Hannes K. Buchholz, Rebecca K. Hylton, Jan Gerit Brandenburg, Andreas Seidel-Morgenstern, Heike Lorenz, Matthias Stein, Sarah L. Price. Thermochemistry of Racemic and Enantiopure Organic Crystals for Predicting Enantiomer Separation. Crystal Growth & Design, 2017, 17(9), 4676–4686. DOI: 10.1021/acs.cgd.7b00582. https://doi.org/10.1021/acs.cgd.7b00582
  6. Lina C. Harfouche, Clément Brandel, Yohann Cartigny, Joop H. ter Horst, Gérard Coquerel, Samuel Petit. Enabling Direct Preferential Crystallization in a Stable Racemic Compound System. Molecular Pharmaceutics, 2019, 16(11), 4670–4676. DOI: 10.1021/acs.molpharmaceut.9b00805. https://doi.org/10.1021/acs.molpharmaceut.9b00805
  7. Simon Clevers, Gérard Coquerel. Kryptoracemic compound hunting and frequency in the Cambridge Structural Database. CrystEngComm, 2020, 22(43), 7407–7419. DOI: 10.1039/d0ce00303d. https://doi.org/10.1039/d0ce00303d
  8. Tharit Lerdwiriyanupap, Giuseppe Belletti, Paul Tinnemans, Ruel Cedeno, Hugo Meekes, Elias Vlieg, Adrian E. Flood. Influence of Ostwald's Rule of Stages in the Deracemization of a Compound Using a Racemic Resolving Agent. Crystal Growth & Design, 2022, 22(2), 1459–1466. DOI: 10.1021/acs.cgd.1c01426. https://doi.org/10.1021/acs.cgd.1c01426
  9. Juan García de la Concepción, Mirian Flores-Jiménez, Louis A. Cuccia, Mark E. Light, Cristóbal Viedma, Pedro Cintas. Revisiting Homochiral versus Heterochiral Interactions through a Long Detective Story of a Useful Azobis-Nitrile and Puzzling Racemate. Crystal Growth & Design, 2023, 23(8), 5719–5733. DOI: 10.1021/acs.cgd.3c00372. https://doi.org/10.1021/acs.cgd.3c00372
  10. Kennedy A. Borchardt-Setter, Lian Yu. Assessing the Potential for Chiral Separation by Crystallization Using Crystal Energies. Crystal Growth & Design, 2023, 23(5), 3615–3622. DOI: 10.1021/acs.cgd.3c00077. https://doi.org/10.1021/acs.cgd.3c00077
  11. Clément Pinetre, Sjoerd W. van Dongen, Clément Brandel, Anne-Sophie Léonard, Maxime D. Charpentier, Valérie Dupray, Kasper Oosterling, Bernard Kaptein, Michel Leeman, Richard M. Kellogg, Joop H. ter Horst, Willem L. Noorduin. Enantiopurity by Directed Evolution of Crystal Stabilities and Nonequilibrium Crystallization. Journal of the American Chemical Society, 2025, 147(10), 8864–8870. DOI: 10.1021/jacs.5c00569. https://doi.org/10.1021/jacs.5c00569 (open access via PubMedCentral 11912332)
  12. Ramamohana Reddy Maddike, Vivek Sharma, Rajasekhar V. S. R. Pullabhotla, Vinod. Engineering Chirality in the Solid State: Co-Crystallization as a Strategy for Chiral Induction and Resolution. Chirality, 2026, 38(5), e70100. DOI: 10.1002/chir.70100. https://doi.org/10.1002/chir.70100
  13. Giuseppe Belletti, Carola Tortora, Indradevi D. Mellema, Paul Tinnemans, Hugo Meekes, Floris P. J. T. Rutjes, Svetlana B. Tsogoeva, Elias Vlieg. Photoracemization-Based Viedma Ripening of a BINOL Derivative. Chemistry – A European Journal, 2020, 26(4), 839–844. DOI: 10.1002/chem.201904382. https://doi.org/10.1002/chem.201904382