When molecules are deposited on a surface, they can self-assemble into a monolayer. Scanning tunneling microscopy (STM) is a powerful instrument to image surfaces with sub-molecular resolution. How does STM work?

A sharp metallic tip is brought very close to a conductive, atomically flat substrate. When a voltage is applied between the tip and the substrate, electrons will tunnel through the barrier. The resulting tunneling current is exponentially dependent on the distance between tip and substrate (height). By monitoring the current and height as it scans across the surface, an image is created and molecules can be seen.

Computational modelling of chiral structures is an essential part of understanding them, and there are several key factors which are controlled in this process:

1. The force Field and the interactions

Force field is the tool used here to get a theoretical insight of chiral assemblies of molecules in solvent and at surfaces. In this representation, the atoms are described as spheres, connected to each other by springs (the chemical bonds). This representation allows us to represent in a simple way the internal energy of the molecule (energy of the bonds, of the angles formed by atoms,…), as well as the interaction energies between molecules, responsible of the formation of supramolecular assemblies, and can be divided in three main families:

• The interactions between molecules are the first involved in the formation of large supramolecular assemblies. They are composed mostly of all interactions available in Nature (Van der Waals interaction, electrostatic and eventually hydrogen bond interactions). The nature of the molecule will determine the relative magnitude of these three components, as well as the impact they have on the global structure. A good example of this relies in the formation of 3D helices upon staking of a certain amount of chiral oligo-phenylenevinylene molecules in solution. In that specific case, the orientation of the molecules is the consequence of repulsion between the chiral centers: Since they’re all directed in the same way, a perfect cofacial stack is not stable, and the molecule tends to form an angle with the value depending on the type of chiral center. With all these angle formed between the molecules, a helical structure is observed, as represented in the cartoon below.
• The interaction with the surface is responsible of the self-assembly we observe in STM images. The interaction between molecule and surfaces, in combination with the interaction between molecules, results in the formation of large domains of molecules. The nature of molecule-surface interaction strongly depends on the nature of surface. For example, if the surface is composed of graphite, the interaction with molecules will mostly occur upon Van Der Waals.
• The molecule-solvent interaction, without being the most important, is ubiquitous. In every case envisioned, with a surface or without a surface, the chiral molecules are mixed in a solvent. This will once again depend on the nature of both chiral molecules and solvent itself.

2. Geometry and energy parameters of chiral structures in solution and at surfaces: how to describe at molecular level.

The use of the modelling tool allows us to extract two different informations on chiral supramolecular assemblies: informations about the “local” geometry of the molecules, as well as the energetics of the system.

2.1  Geometry

The deep analysis of geometries is a typical example of explanation brought by modelling. Since we’re dealing with atom-by-atom representation of the whole molecules, we can have a look at “what’s inside” and to explain the properties observed by experimental ways. The “nucleation process” of a chiral helix is a good example of this. In order to form a helix, the stack of chiral molecules must achieve a certain size (the so-called nucleation size). Once it is obtained, the helicity appears. If we measure the angle formed between the molecules and of we look their evolution in-time with modelling tools, we are able to estimate the moment where a clear helicity appears, it is to say the size of the nucleus. A complementary approach to this consists in the interaction of these stacks with light, that allows us to draw similar conclusions: one observes the typical signature of a specific helicity that reinforces with the size of the molecule stack.

2.2  Energies

A careful study of energies in simulations gives an excellent way to understand the very mechanisms underlying the formation of chiral assemblies, as well as the selectivity of an enantiomer. In the specific case of separation racemic mixtures by an enantiomeric compound, one expects to get a “hole-key” process, with a specific recognition of the two elements with the same chirality. For example, if the element is separating compound is a R enantiomer, we expect a specific recognition of the R element in the racemic. This parameter is driven by interactions, and thus appears in the expression of energies. A closer look at the combination of enantiopure acid with two amine enantiomers, for example, revealed that the most stable case corresponds to the combination of two identical enantiomers.

A clean surface is a blank canvas to the chiral artists at Resolve. We add chiral art to that canvas by painting it with molecules. We use two techniques to stick, or adsorb, molecules to a clean surface. The first is evaporation under vacuum conditions, and the second is deposition from solution. In the former a sample of molecules is heated in vacuum. Some of the molecules evaporate individually and travel to the surface where they stick. Subsequent processing steps can change the manner in which the molecules adhere to and organize upon the surface. To adsorb molecules from solution a droplet containing molecules dissolved in a solvent is placed upon the clean surface. The molecules diffuse to the surface and adsorb.

Once adsorbed onto the surface, the molecules can induce a surprisingly complex, and hierarchical expression of chirality. In fact, it is much easier to induce chirality at a surface and objects which are inherently achiral in three-dimensions can become chiral upon confinement at an interface. At the smallest scale, a chiral molecule adsorbed to a surface will retain its chirality. A molecule can, on adsorption, express chirality in its adsorption footprint, the manner in which it binds to the surface. At a larger scale many molecules may organize into regular patterns which themselves may be chiral leading to a macroscopic chiral surface.

In 1998, a group of Dutch researchers tried to make the screening process for resolution by diastereomeric salt formation faster by adding stochiometric amounts of several resolving agents as a mixture to the racemate. They soon discovered that random combinations of resolving agents did not give good results. Only very insoluble salts can be selected using this method. However, when the researchers used structurally related and homochiral resolving agents (family members) the outcome was different. Often, the combination of these resolving agents gave higher ee’s than with each of these resolving agents independently. Moreover, the chance of obtaining solid salts with significant diastereomeric excesses was increased from 20–30% to 90–95%.

Reverse Dutch Resolution has been reported where family members of a racemate have been resolved simultaneously with one resolving agent. Without the addition of one of the family members, the resolution fails.

The reasons for the high success rate of Dutch Resolution is believed to be:

Choice of the best resolving agent/racemate combination. With three resolving agents and one racemate, the least soluble combination of diastereomers will start to crystallize, hereby reducing the chance of encountering a salt that will not form crystalline salts.

Solid solution behaviour of the family members. Solid solution behaviour means that the crystal lattice does not distinguish much between the several family members that can fit inside the crystal lattice of the salt that is precipitating. Hence, the composition of the crystal depends largely on the composition of the surrounding solution. Furthermore, the solubility of such a solid solution is lower than for each separate salt combined.

Peachey-Pope type resolution. In a Peachey-Pope resolution, instead of one equivalent of, say, mandelic acid as resolving agent, one-half equivalent of mandelic acid is used and supplemented with one-half equivalent of an achiral (cheap) acid like hydrochloric acid to make the system neutral. The achiral supplement should give very soluble salts with the racemate so these will not crystallize and ruin the resolution. The less soluble salt will start to crystallize and will consume most of the resolving agent thus leaving only small amounts of resolving agent for the more soluble diastereomer which, in an ideal case, will not crystallize. The same principle applies to Dutch resolution. When three resolving agents are used, usually, one of these is incorporated the most in the least soluble diastereomer. The concentration of this resolving agent in the solution is subsequently lowered and thus the more soluble diastereomer of this resolving agent cannot crystallize.

Nucleation inhibition. When a family of three resolving agents is used, sometimes only two are incorporated in the crystal lattice. It was found that if a resolving agent is not incorporated in the crystal lattice, this does not mean this compound can be left out. Small amounts of compounds that resemble the resolving agent or racemate can inhibit the nucleation of the more soluble (unwanted) diastereomer.

Nucleation inhibition has been further investigated and was found to be very effective for the improvement of a classic resolution by diastereomeric salt formation. By addition of only a few percent of a structurally related additive, significant improvements were found. This type of resolution was named Second Generation Dutch Resolution