Highlights from the Faraday Discussion: Halogen Bonding in Supramolecular and Solid State Chemistry, July 10–12th 2017, Ottawa, Canada

The first Faraday Discussion on halogen bonding was held in the beautiful capital city of Canada, Ottawa, from 10^th to 12^th July, 2017. Although the delegates arrived at the conference venue on the edge of the lovely University of Ottawa campus in the rain (Fig. 1), the mood at registration and the opening lunch was not dampened by the inclement weather outside. This cheerful atmosphere continued throughout the meeting: although the halogen-bonding community is extremely diverse considering the multi-disciplinary nature of the approaches used to understand the nature and role of halogen bonding in supramolecular chemistry and the solid state, it is nevertheless surprisingly tight-knit. This led to a strong feeling of collegiality at the meeting, characterised by lively discussion.

Fig. 1 The exterior of the Desmarais building, part of the University of Ottawa, Canada, where the conference was hosted. (Photograph courtesy of Dr. Angshuman Roy Choudhury.)

After the opening lunch and a welcome from the conference chair David Bryce (University of Ottawa, Canada), the formal part of the meeting started off with a bang with the thought-provoking introductory lecture by Tim Clark (University of Erlangen-Nürnberg, Germany) on ‘Halogen Bonding and σ-holes’ (DOI: 10.1039/c7fd00058h). Prof. Clark immediately piqued the interest of the audience by giving a historical background to Occam's razor. This principle, also known as lex parsimoniae, is that the simplest explanation is the most likely one. Although Prof. Clark pointed out that this is not true for biology, he was able to use the lex parsimoniae principle as a basis for his approach to bonding theory. To this end, he demonstrated that only Coulomb attractions, polarisation and dispersion are needed to describe weak intermolecular bonding. He showed, for instance, that charge transfer, which is only meaningful at infinite separation and requires the concept of orbitals or boundaries between atoms, can be explained in terms of polarisation since this yields exactly the same effect. In addition, he utilised the Hellman–Feynman theorem to show that even dispersion is not a physical effect and highlighted the role of Coulombic interactions through polar flattening and the presence of σ-holes. He ended his presentation with several examples, focussing particularly on using electrostatic models to describe the role of polarisation in the formation of a wide variety of halogen bonds and other σ-hole interactions, including the deceptively simple case of Br₂ interacting with Ar.

Session 1: Computational approaches and σ-hole interactions

The first session on computational approaches and σ-hole interactions (chaired by Catharine Esterhuysen, Stellenbosch University, South Africa) was kicked off by Janet Del Bene (Youngstown State University, USA) describing her work on halogen bonds within ternary complexes (DOI: 10.1039/c7fd00048k). She was able to show computationally that increasing the strength of one halogen bond within the ternary system results in a concomitant increase in the strength of a second halogen bond, to the extent that in H₂XP:ClF:ClF systems (where X = F, Cl, H and NC) the resultant strong halogen bonds lead to Cl transfer to yield the (H₂XP–Cl)⁺:⁻(F:ClF) ion pair. On the other hand, the halogen bonds within H₂XP:ClF:ClH are weaker owing to the less prominent electron-withdrawing ability of H relative to F. It was pointed out during the discussion session that the terminal ClH molecule could also have engaged in a hydrogen bond, leading to a discussion comparing hydrogen bonding and halogen bonding in such systems. In particular, the close resemblance between the effect noted by Dr. Del Bene and the cooperativity effect known for hydrogen bonding was a source of subsequent discussion.

The second speaker was Kevin Riley (Xavier University of Louisiana, USA), who focussed on halogen bonds involving cationic halogen bond donors (DOI: 10.1039/c7fd00106a). He showed computationally that in H₃N–C≡C–Br⁺⋯Cl⁻ the most stabilising conformation involves a Cl⁻⋯Br–C angle of 180°, as expected considering the highly directional nature of halogen bonding, with the interaction energy decreasing as the Cl⁻⋯Br–C angle decreases. Surprisingly, however, a close approach of the Cl at a Cl⁻⋯Br–C angle of 90° proved to be more stable than at 120°. This is due to enhanced electrostatic attraction as a result of secondary interactions between the Cl⁻ and the ammonium group that are only possible at angles less than 120°. In order to estimate the strength of the interaction within a crystal, the system was modelled within a solvent using a polarisable continuum model. The applicability of this approach was the source of active discussion, with Prof. Riley pointing out that even though a solvent medium is not an accurate description of the electrostatic environment in a crystal it is more accurate than in vacuo calculations.

This discussion led neatly into the third paper of the session on the interactions between clathrate hydrate cages and their dihalogen guests (DOI: 10.1039/c7fd00064b) presented by Saman Alavi (University of Ottawa, Canada). He showed that, in order to obtain an accurate computational description of the halogen bonding involved, it was necessary to model the clathrate hydrates' interaction with a surrounding shell of water molecules through hydrogen bonding. This prevented over-binding of the dihalogen molecule with lone pairs on the cage water molecules.

The very lively discussion that concluded the first half of the session raised, amongst others, the issue that experimentalists are confused about choosing from the large range of models that theoreticians use. Added to this is the problem that some journals require computational results to be reported, leading to the publication of poor quality data that are not sensible or reliable. The intensity of the discussion is captured in Fig. 2.

Fig. 2 General discussion of the first half of the first session. (From left to right: Saman Alavi, Janet Del Bene, Kevin Riley and Tim Clark. Photograph courtesy of Dr. Angshuman Roy Choudhury.)

Seiji Tsuzuki (National Institute of Advanced Industrial Science and Technology, Japan) also gave a short presentation on his view that long-range interactions (electrostatic and dispersion) are the driving force behind halogen bonds. Although Dr. Del Bene agreed, pointing out that the presence of covalent bonding does not necessarily mean that it is the driving force, Pierre Kennepohl (University of British Columbia, Canada) noted that covalency can be used to explain chemical properties of a system.

After the tea break, Prof. Kennepohl was able to expand on his statement mentioned above, by showing that the halogen bonds within I₂X⁻ and I₄X⁻ species (where X = Cl or Br) involve a high degree of covalency (charge transfer), based on his group's work using X-ray Absorption Scattering (XAS) and Density Functional Theory (DFT) calculations (DOI: 10.1039/c7fd00076f). He concluded that the halogen bonds in polyhalides can be seen as being analogous to coordination bonds, and even suggested that they could be indicated with an arrow, rather than the three dots usually used.

Alison Edwards (Australian Centre for Neutron Scattering, Australia) then addressed the question of ‘Intermolecular interactions in molecular crystals: what's in a name?’ (DOI: 10.1039/c7fd00072c). She showed that visualisation of various energetic components of a series of intermolecular interactions not only highlights their common origin, namely redistribution of electron density, but also gives insight into the properties of crystals. There was much discussion regarding whether this approach could be used for prediction; as Dr. Edwards pointed out, the holistic approach suggested in their paper should be very helpful, even though there are many examples of multiple solid-state structures which complicates matters.

The final speaker of the session, Peter Politzer (University of New Orleans, USA), continued with the theme of the nature of halogen bonding interactions (DOI: 10.1039/c7fd00062f). In his paper on close contacts in crystals he pointed out that close contacts within the sum of the van der Waals' radii do not necessarily indicate non-covalent interactions. Instead, utilising the concept of the σ-hole as a basis, he echoed Tim Clark's words in the introductory lecture that any attraction present can be explained by the Hellman–Feynman theorem. Prof. Politzer's paper served as the starting point for an animated debate regarding the use and usefulness of electrostatic potentials in the understanding of a variety of different interactions before a well-deserved tea break.

The day's formalities were ended with a good-humoured and relaxed poster session. The extremely high standard of the posters ensured lively conversations, but left the poster-prize judges with a difficult decision. Notably, the diverse subjects covered during the poster session, from mechanochemical investigations of halogen-bond formation to crystallography, emphasised the broad spectrum of interest that the halogen-bonding community has harnessed.

Session 2: Beyond the halogen bond

The second day started with Session 2, chaired by David Bryce. Scott Southern (University of Ottawa, Canada) launched the session on the topic of ²⁰⁷Pb solid-state NMR investigations of lead metal–organic frameworks, featuring lead tetrel bonds (DOI: 10.1039/c7fd00087a). Several beautiful examples of lead-centred compounds were presented, with the intention of characterising the tetrel bonding environment using ²⁰⁷Pb solid-state NMR. A survey of literature ²⁰⁷Pb solid-state NMR data categorised the NMR responses according to two coordination geometries surrounding the lead centres: hemidirected and holodirected. The experimental NMR results yielded insight into the nature of the short contacts to the open faces of the lead centers, supporting the observation of the tetrel bonds.

It was later discussed that the “tetrel group” (group 14 of the periodic table) consists mostly of spin-1/2 nuclei, with the exception of Germanium-73, which is quadrupolar (spin 9/2). Solution and solid-state NMR investigations of the tetrel bond are therefore readily accessible, with few examples presented in the literature.¹ In the case of solid-state NMR, these spin-1/2 nuclides can yield large spectral widths in asymmetric environments, requiring high resolution techniques or signal enhancement, such as magic-angle spinning and the Carr–Purcell–Meiboom–Gill method. In this study, these techniques were employed in the acquisition of the ²⁰⁷Pb spectra, setting a standard for solid-state NMR investigations of tetrel bonding.

Following this was a presentation by Ignacio Vargas-Baca (McMaster University, Canada), discussing the self-assembly of iso-tellurazole N-oxides, forming macrocycles through Te⋯O chalcogen bonds (DOI: 10.1039/c7fd00075h). These macrocycles were studied by solution NMR in order to investigate under which condition the Te⋯O chalcogen bonds would be disrupted, for instance, through the addition of Lewis bases and acids. Following the addition of BR₃ (R = Ph, F) boranes, the products were investigated by single crystal X-ray diffraction, revealing characteristic chalcogen bonds involving tellurium. A computational analysis was carried out, offering further insights into the strength and nature of the chalcogen bond.

Following the general discussion, morning tea was served in the lobby, with conference attendees helping themselves to fresh coffee, tea, and baked goods. A positive mood flowed through the conference, with the exchange of both ideas, and pleasantries. A picture capturing the moment (Fig. 3) shows a group socialising before the start of the second half of the session.

Fig. 3 Conference attendees enjoy a hot beverage during the morning tea break. (From left to right: Lee Brammer, Tim Clark (behind), Tomislav Friščić, Mate Erdelyi, Bill Pennington, Christer Aakeröy, Marta Mosquera, David Bryce and Tayur Guru Row. Photograph courtesy of Dr. Angshuman Roy Choudhury.)

The second half of the session was initiated by Tayur Guru Row (Indian Institute of Science, India) on the topic of solid solutions and the effect of halogen bonding on their packing preferences (DOI: 10.1039/c7fd00084g). Two isomers, 4-bromo-2-chlorobenzoic acid and 2-bromo-4-chlorobenzoic acid, were investigated by single crystal X-ray diffraction, revealing triangular motifs held together by halogen⋯halogen contacts. The crystal structures of solid solutions were then investigated by varying the stoichiometry of both components, in order to establish the role of halogen bonding in the crystal packing. The solid solutions were then studied by differential scanning calorimetry, in order to gain information on their thermal stability.

The final presentation of the second session was given by Steve Scheiner (Utah State University, USA), on a computational study comparing halide receptors utilising a series of non-covalent interactions (DOI: 10.1039/c7fd00043j). The halide receptors were investigated in terms of both the acceptor halide, as well as different binding atoms. In total, all four halides were investigated (F⁻, Cl⁻, Br⁻ and I⁻) as well as the comparison between the hydrogen bond, the halogen bond, the pnictogen bond, the chalcogen bond, and finally the tetrel bond. Investigations suggest that F⁻ was the strongest binding halide, while the Sn tetrel bond fared best for selectively binding the F⁻ anion, with a six-fold enhancement over the hydrogen bond receptor.

Session 3: The halogen bond in solution

The afternoon session, chaired by Kevin Riley, focused on the study of halogen bonding in solution, which is a field that has only recently developed, but is currently expanding. During this session, several aspects of this field were addressed, from the impact of the halogen bond on reactivity to supramolecular assembly and recognition, with an overview of the different tools that can be used to probe this specific interaction.

Paul D. Beer (University of Oxford, UK) started the session off with the presentation of his paper on the study of cationic rotaxanes designed for specific sensing of halides through the formation of halogen bonds (DOI: 10.1039/c7fd00077d). The binding ability with anions within the cavities of three different rotaxanes, including a particular one with a rhenium metal complexed with the macrocycle, was evaluated by ¹H NMR spectroscopy. His results showed halide selectivity over oxoanions in organic–aqueous solvent media, with specificity for particular halides depending on the structure of the rotaxane.

The second speaker in the session, Marta E. G. Mosquera (Universidad de Alcalá, Spain), expounded on an extensive study on the different solid networks obtained upon addition of elemental halogen (I₂ or Br₂) to Ru(X)₂(CNR)₄ complexes (X being a chloride or a bromide and CNR an alkyl or aryl isocyanide, DOI: 10.1039/c7fd00079k). She addressed the role of halogen bonding on the unexpected reactivity of these ruthenium(II) complexes in solution where substitution of the X ligand takes place despite no halide being present. The combination of crystallographic analysis, Raman spectroscopy and ¹H and ¹³C NMR spectroscopy helped her to build a solid argument regarding the reaction pathway.

Mark S. Taylor (University of Toronto, Canada) then presented his work on the formation of multicompartmental micelles using halogen bonding to promote the supramolecular assembly of complementary polymers (DOI: 10.1039/c7fd00111h). Mixing of specifically designed triblock terpolymers possessing a halogen bond receptor block with a halogen bond donor polymer yields nanostructures able to collapse to multicompartmental micelles upon transfer from organic solvent to water. This particular behaviour, as well as the wide variety of nanostructures and aggregates obtained, have been characterised using transmission electron microscopy and electron energy loss spectroscopy.

In his presentation after the tea break, Mathieu Branca (Université Paris Diderot, France) showed an interesting combination of electrochemical and NMR studies to probe and characterise the effect of different possible interactions upon addition of anions on the reduction potentials of viologen derivatives (DOI: 10.1039/c7fd00082k). Although the effect of halogen bonding on the reduction potentials of a halogen bond donor phenylviologen is clear, hydrogen bonding and ion pairing involving the chloride occurring on the viologen core also have a strong influence. The discussion that followed mainly focused on the possibility that π-anion interactions may be responsible for the particular behavior of pentafluorophenylviologen.

The next speaker, Sergiy V. Rosokha (Ball State University, USA), presented his study on electron transfer reactions of halogenated electrophiles (DOI: 10.1039/c7fd00074j). The comparison of experimental data with calculations allowed him to show the limitation of the outer-sphere dissociative electron-transfer model to predict the rate constants when a transient halogen bond formed between the electron donor and the acceptor. He thus demonstrated that estimation of electronic coupling elements in such systems, via the Mulliken–Hush formula, was necessary to obtain consistent rate constants because the formation of a halogen bond substantially decreases the electron transfer barrier.

The last speaker of the day, Mate Erdelyi (University of Gothenburg, Sweden), described his article on using ¹⁵N NMR spectroscopy to probe very weak halogen-bonding complexes (DOI: 10.1039/c7fd00107j). Although DFT-level calculations indicate that ¹⁵N NMR spectroscopy is a promising tool to probe this interaction, solution NMR studies give significant results only for the strongest interactions between pyridine and diversely para-substituted iodobenzenes. Comparison of the systems obtained upon halogen-bond formation with their hydrogen-bonded counterparts using phenols instead of iodobenzenes, showed that chemical shifts were an order of magnitude larger in the case of hydrogen bonding than for halogen bonding.

Although there are few researchers that specifically work on halogen bonding in solution, the discussions that followed each presentation were very lively, with strong participation even from the longtime specialists in halogen bonding from a theoretical or solid state perspective.

The day ended with the conference dinner in the beautiful Hugette Labelle Hall. The first highlight of the evening was the presentation of the poster prizes to Chantal Mustoe (University of British Columbia, Canada) and Patrick Szell (University of Ottawa, Canada) by Eleanor Campbell, president of the Faraday division of the Royal Society of Chemistry (Fig. 4). Prof. Campbell then gave an entertaining dinner talk, before guiding us through the second highlight, namely ‘the Loving Cup ceremony’, a Faraday Discussions tradition that led to much hilarity.

Fig. 4 Poster prize winners (receiving their prizes from Eleanor Campbell, inset): Chantal Mustoe (left; X-ray Absorption Spectroscopy as a Probe for Weak Interactions) and Patrick Szell (right; Multinuclear Solid-State Magnetic Resonance of 2-/3-iodoethynylpyridine and Their Hydrohalide Salts: A Direct Comparison between the Halogen Bond and the Hydrogen Bond). (Photographs courtesy of Catharine Esterhuysen and the Royal Society of Chemistry.)

Session 4: Solid-state chemistry and applications

The fourth session of the conference took place on the Wednesday morning and was chaired by William Pennington (Clemson University, USA). This last session was dedicated to the applications of σ-hole bonds in the solid state, as well as experimental studies involving diffraction techniques and solid-state synthesis.

During the first part of the session, Christer Aakeröy (Kansas State University, USA) presented his paper on competition and selectivity in supramolecular synthesis (DOI: 10.1039/c7fd00080d). The presence of different chemical functionalities in the starting synthons often leads to competition among binding sites, hindering the construction of supramolecules with specific architectures and metrics. To shed light on this issue, the impact of electrostatics and geometry on supramolecular structures was explored through the systematic cocrystallisation of three isomers of 1-(pyridylmethyl)-2,2′-biimidazole with selected hydrogen- and halogen-bond donors. Prof. Aakeröy showed that electrostatics played a decisive role in determining supramolecular selectivity, and electrostatically and geometrically favorable two-point interactions deeply influenced the final structures. Halogen-bonded complexes showed much more consistent and predictable structural features compared to the analogous hydrogen-bonded ones, and were all based on the simultaneous presence of self-complementary N–H⋯N/N⋯H–N motifs and I⋯N(pyridine) halogen bonds. Overall, this work confirmed that concurrent hydrogen and halogen bonds are useful tools for refining supramolecular synthetic strategies.

The following speaker was Marc Fourmigué (Université de Rennes & CNRS; France), who reported a comparative structural and computational study on two halogen-bonded bimolecular adducts formed by strong halogen bond donors – namely, N-bromo- and N-iodosaccharin – with 4-picoline as the acceptor (DOI: 10.1039/c7fd00067g). In contrast to what was observed for weaker halogen bonds, the bromo-adduct exhibited a stronger halogen bond than its iodo-analogue. Both X-ray crystal structures and topological analyses of electron density confirmed that in the bromo-derivative, there was a more pronounced displacement of the halogen atom toward the picoline, affording a more “ionic” structure of the final complex.

After the morning tea break, the second part of the session was focused on some potential applications of halogen bonds in materials science, with examples belonging to both the soft matter and solid-state fields. Regarding the former topic, Valentina Dichiarante (Politecnico di Milano, Italy) showed that halogen bonding can efficiently drive the formation of photoresponsive ionic liquid crystals, starting from non-mesomorphic building blocks (DOI: 10.1039/c7fd00120g). The self-assembly of azobenzenes bearing an iodo-tetrafluoro-phenyl ring and methyl-alkyl-imidazolium iodides actually led to supramolecular liquid crystals with good thermal stability. The presence of photoactive azo-groups allowed light-induced phase transitions from smectic A liquid crystalline phases to isotropic liquids, whereas the backward transition was achieved by cooling. Dr. Dichiarante suggested that the combination of photoresponsivity and the ionic nature of these complexes may offer a new route towards light-induced control over ion transport and conductance.

Another interesting class of materials is represented by gels, which were the main topic of the paper presented by Gareth Lloyd (Heriot-Watt University, UK), concerning the effect of halogens on the gelation properties of phenylalanine (DOI: 10.1039/c7fd00108h). Phenylalanine (Phe) is an essential amino acid for human health and industrial synthesis, and is known to form gels both in water and in dimethylsulfoxide. Gelation performance and rheological properties of a series of Phe derivatives bearing different halogen atoms on the aromatic ring were screened in both solvents. The resultant hydrogels were opaque, crystalline in nature, and easily collapsed under mechanical manipulation. On the other hand, the organogels were transparent, not crystalline, and difficult to compress. The presence of I or Br in the para-position of the aromatic ring, while increasing the strength of halogen–halogen interactions, led to slower exchange dynamics at the gel/solution interfaces. This resulted in metastable gels in the case of 4-bromo-Phe, and even in the loss of gelation ability with 4-iodo-Phe.

The last speaker was Tomislav Friščić (McGill University, Canada), who demonstrated that halogen-bonding cocrystallisation is a viable strategy to obtain supramolecular structures with dichroic behaviour, not only starting from organic molecules, but also from metal–organic materials (DOI: 10.1039/c7fd00114b). The assembly of azobenzene chromophores with dicyanoaurate ions (linear ditopic metal–organic acceptors) led to a parallel alignment of all halogen-bonded chains in the crystal packing, with resulting dichroism. A liquid crown ether (15-crown-5) was added to enhance the solubility of KAu(CN)₂ through complexation of the potassium cation. Three 15-crown-5 molecules were incorporated into the asymmetric unit of the crystal structure, resulting in a rather unusual ‘four-component solid’, that can be considered a cocrystal (halogen-bonded chains), a salt (combination of ionic species) and a solvate (inclusion of a liquid in the crystal lattice) at the same time.

The final general discussion dealt with the role that computed X-ray crystal structures may exert in directing the design of supramolecular functional materials. A rather animated debate arose among the delegates, pointing out that predicted structures surely represent a fundamental tool for understanding experimental results, although several possible biases might affect their reliability. Prof. Friščić, in particular, expressed the wish that the scientific community could become more open and collaborative in reporting research failures, and not only successful achievements. He also suggested that a broader exchange of ideas and opinions among experimental scientists and theoreticians should be promoted, in order to better understand the reasons behind such disappointing results and to find possible solutions.

Concluding remarks

The concluding remarks lecture “Halogen Bonding and Beyond” was presented by Lee Brammer (University of Sheffield, UK), who recapitulated the meeting by giving a brief overview of all the topics covered during the Discussion. In particular, he offered an interesting and original point of view on the conference themes, and proposed that we classify the papers into three main thematic areas:

(i) Fundamentals. Keeping in mind the lex parsimoniae principle cited by Prof. Clark in his introductory lecture, several studies aimed to assess the role of charge transfer, polarization and covalency in these supramolecular interactions. Important theoretical advances were reported regarding the competition between XB and HB, the nature of σ-holes, the influence of close contacts and noncovalent intermolecular interactions on the self-assembly of these complexes.

(ii) Characterisation. Different techniques were successfully applied to the characterisation of halogen-bonded systems, among which UV-visible spectroscopy, Nuclear Magnetic Resonance (both in solution and in the solid state) and electrochemistry played a key role.

(iii) Applications. A series of papers were dedicated to studying the potential of halogen bonding and its related interactions for applied science. A broad range of applications was shown, spanning halide–anion binding, synthesis and reactivity of supramolecular complexes and cocrystals and the preparation of multicompartmental micelles, gels and photoresponsive ionic liquid crystals to metal–organic solids with dichroic properties.

Prof. Brammer concluded his lecture highlighting the importance of this unique Faraday Discussion in unveiling the deep connection between classical halogen bonds and a much broader array of σ-hole interactions. The well-established principle that “blue bits interact with red bits and vice versa” may lead to a completely new awareness of the mechanisms involved in s- and p-block chemistry.

Finally, Anthony Legon (University of Bristol, UK) closed the meeting by acknowledging the organizers, the RSC staff and all the participants (Fig. 5). Then, the delegates had a last chance to discuss the talks and posters during the final lunch.

Fig. 5 Group photo of the conference attendees. Scott Zablotny (front left, University of Ottawa) and Estelle Caron-Poulin (front right, University of Ottawa), stewards, along with the conference organizers Jeremy Allen and Marie Cote (front left, RSC), were integral in the smooth operation of the conference. (Photograph courtesy of the Royal Society of Chemistry.)

We thank Dr. Angshuman Roy Choudhury for providing us with photos from the conference.

References

1. S. A. Southern and D. L. Bryce, J. Phys. Chem. A, 2015, 119, 11891–11899.

---