Electrophilic and nucleophilic gas phase reactivity of the Janus cluster-based anions [{Mo₆Cl₈}Cl₅□]⁻ (□ = lacuna)

Abstract

The lacunary monocharged anion [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ presents concomitantly a strongly electrophilic site and a nucleophilic one. This Janus character in terms of reactivity is confirmed by its gas phase reaction with [Br₆Cs₄K]⁻ to form [{Mo₆Clⁱ₈}Cl^a₅Br^a]²⁻ and by its unusual self-reactivity leading to [{Mo₆Clⁱ₈}Cl^a₆]²⁻ dianions.

---

Cluster-based ternary halides of general formula A₂[{Mo₆Xⁱ₈}X^a₆] (A = Cs⁺, Rb⁺, K⁺; X = Cl, Br or I; inner ligand “i”, apical ligand “a”) ([{Mo₆Xⁱ₈}X^a₆]²⁻ depicted in Scheme 1) are prepared by solid-state chemistry at high temperature. Contrary to many other inorganic solid-state compounds, the A₂[{Mo₆Xⁱ₈}X^a₆] halides are soluble in acidified aqueous solution and in common organic solvents, which allows their facilitated integration in functional materials. In all these materials, the [{Mo₆Xⁱ₈}X^a₆]²⁻ cluster units or their derivatives (obtained by ligand substitutions) are the active species, which experience during integration bond energy activations, nucleation processes, chemical reactions or electronic excitations and further processes. These numerous and various properties (photoluminescence, (photo-)(electro-) catalysis, light harvesting properties) open the way for a wide range of potential applications in the field of display and lighting technologies, sustainable energy and health.^1,2 This is thus of prime interest to study the intrinsic properties and reactivities of discrete cluster units in order to perform the rational design and elaboration of materials.^3–5 Some gas phase studies of [{Mo₆Xⁱ₈}X^a₆]²⁻ dianions (X = Cl, Br and I) have been recently conducted using collision-induced dissociation (CID) mass spectrometry (MS) measurements and quantum chemical calculations.⁶ In particular, the fragmentation pathways initiated by the loss of an X⁻ anion, and the sequential eliminations of MoX₂ and X˙ radicals have been described. The propensity of fragment ions to form O₂ or H₂O adducts has also been highlighted.

Scheme 1 Description of the ESI-MS fragmentation process of [{Mo₆Clⁱ₈}Cl^a₆]²⁻ leading to [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻; reactivity of [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ with [Cs₄KBr₆]⁻; [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ self-reactivity.

In the present work, we demonstrate that the weakly nucleophilic monocharged [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ (□ = lacuna) anions prepared in the gas phase from [{Mo₆Clⁱ₈}Cl^a₆]²⁻ by a Cl^a− ligand loss represent one of the few examples of anionic electrophiles. Surprisingly, we observed that these negatively-charged electrophilic ions readily react in the gas phase with {Cs₄KBr₆}⁻ anions by bromine abstraction, despite the coulombic repulsion. Quantum chemical studies of the isolated [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ have been combined with MS experiments to secure the interpretation. These results are compared and discussed with those reported for [B₁₂L₁₁]⁻ (L = F, Cl, I or CN) superelectrophilic anions.^7–11

Geometry optimizations and the subsequent computational studies were performed at the DFT level (see ESI† for details). The Mo–Cl^a heterolytic bond dissociation energies (ΔE_bond) were calculated for [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ and [{Mo₆Clⁱ₈}Cl^a₆]²⁻ cluster units. The Mo–Cl^a ΔE_bond values in [{Mo₆Clⁱ₈}Cl^a₆]²⁻ ([{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ + Cl⁻) and in [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ (trans-[{Mo₆Clⁱ₈}Cl^a₄□^a₂]⁻ + Cl⁻) were determined by an energy decomposition analysis, as reported in Table 1. This highlights the iono-covalent character of the Mo–Cl^a bonds, for which the main component is electrostatic. The latter is much stronger in [{Mo₆Clⁱ₈}Cl^a₆]²⁻ than in [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻. It is apparently counterintuitive since the two fragments resulting from the bond cleavage in [{Mo₆Clⁱ₈}Cl^a₆]²⁻, i.e. Cl⁻ and [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻, are negatively charged and should generate electrostatic repulsions when interacting. Mapping the computed electrostatic potential of [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ reveals its non-uniform distribution with alternation of positive and negative potentials on the Mo and Cl atoms, respectively (Fig. 1). Thus, if a negatively-charged reactant is approaching along the dipole moment axis of [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻, an electrostatic attraction will occur since the Mo atom bearing the lacuna is strongly positively charged even though the molecule is globally negatively-charged. Interestingly, the most negative regions are located around the five apical ligands (Cl^a).

Table 1 Calculated heterolytic Mo–Cl^a bonding dissociation energies ΔE_bond in [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ and [{Mo₆Clⁱ₈}Cl^a₆]²⁻ (in eV). The energy decomposition is given in terms of fragment distorsion energy (ΔE_prep), Pauli repulsion (ΔE_Pauli), electrostatic (ΔE_elstat), orbital interactions (ΔE_orb), dispersion (ΔE_disp), and basis set superposition error correction (ΔE_BSSE).^12–17 Zero-point vibrational corrections are not included

	[{Mo6Cl^i8}Cl^a5□^a]^−	[{Mo6Cl^i8}Cl^a6]^2−
ΔEprep	0.282	0.083
ΔEPauli	4.163	5.222
ΔEelstat	−2.303	−5.357
ΔEorb	−2.696	−3.291
ΔEdisp	−0.375	−0.395
ΔEBSSE	0.030	0.030
ΔEbond (in kJ mol^−1)	−0.898 (−86.6)	−3.708 (−357.7)

---

Fig. 1 Left: Fukui functions for electrophilic (f+) and nucleophilic (f−) attack (±0.001 au) of [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻. Right: Electrostatic potential (in Hartree) mapped onto the isodensity (±0.03 e Bohr⁻³); representation of the dipole moment [small mu, Greek, vector] (6.52 Debye) by an arrow (origin = Mo₆ centre).

The visualization of the Fukui functions for electrophilic (f⁺) and nucleophilic (f⁻) attacks represented in Fig. 1 unambiguously prove the Janus character of [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻. Indeed, it is sensitive to electrophilic attack on the Cl ligand in the trans position to the lacuna of Cl⁻, and at the same time the Mo atom without an apical ligand is highly sensitive to nucleophilic attack.¹⁸ It has to be emphasized that these two reactive sites are positioned on the dipole moment axis. The calculated Gibbs free energy of the reaction 2 × [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ → [{Mo₆Clⁱ₈}Cl^a₄□^a₂]⁰ + [{Mo₆Clⁱ₈}Cl^a₆]²⁻ is 263 kJ mol⁻¹ at 298 K, which is thermodynamically accessible considering the kinetic energies of the ions subjected to an applied electric field (as in the MS experiment described below). It has to be noted that the calculations of bonding and total energies for the [Br₆Cs₄K]⁻ ion could not be considered in the present study since structural information is lacking. Several tens of possible geometrical arrangements could be envisioned and would be the subject of a computational study on their own.

The [{Mo₆Clⁱ₈}Cl^a₆]²⁻ dianion has been identified by negative ion mode electrospray ionization high resolution MS (ESI-HRMS) analysis of a Cs₂[{Mo₆Clⁱ₈}Cl^a₆] acetonitrile solution as proved by the high resolution mass spectrum given in Fig. S1 (ESI†). Indeed, the latter shows an isotopic distribution centered at m/z 535.5 that perfectly matches the theoretical isotopic distribution of a Mo₆Cl₁₄ dianion (see ESI† for details). A second isotopic pattern centered at m/z 1036.0 is observed. It corresponds to the formation of monocharged [Mo₆Cl₁₃]⁻ ions that result from the loss of a Cl⁻ ligand from [{Mo₆Clⁱ₈}Cl^a₆]²⁻. This in-source collision induced dissociation (CID) occurs in the desolvation interface that separates the atmospheric pressure of the ESI-MS source from the analyzer vacuum of the Waters Synapt G2-Si mass spectrometer. This phenomenon has already been reported for the [{Mo₆Xⁱ₈}X^a₆]²⁻ ions.⁶

Second-generation fragment ions are further produced in the course of CID experiments from the in-source-generated [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ fragment ions by selecting the whole isotopic pattern centered at m/z 1036. Surprisingly, as revealed by the CID mass spectrum shown in Fig. 2, [{Mo₆Clⁱ₈}Cl^a₆]²⁻ ions are produced and seem to arise from a chlorine ion addition to the mass-selected [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻.

Fig. 2 Negative ESI-MS-MS spectrum of the [Mo₆Cl₁₃]⁻ ion selected in a mass range encompassing an experimental isotopic pattern centered at m/z 1036. Signal magnification of 10 times in m/z 500–800.

The formation of these [{Mo₆Clⁱ₈}Cl^a₆]²⁻ ions can only be explained based on a bimolecular interaction between two [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ ions concomitantly present in the T-wave (traveling wave) collision cell, as described in Scheme 1.

To confirm this, complementary MS experiments were performed. First, a solution of Cs₂[{Mo₆Clⁱ₈}Cl^a₆] and cesium bromide, CsBr, in acetonitrile was submitted to ESI-MS analysis (see ESI† for details). A full-scan ESI-MS spectrum of the solution in negative ion mode is presented in Fig. 3. Beside the most abundant [{Mo₆Clⁱ₈}Cl^a₆□^a]²⁻ ions, additional anions were detected. m/z 718.4 and 930.2 ions are characteristic of adducts produced in the gas-phase by ESI-MS of cesium bromide in solution. More interesting is the detection of abundant m/z 558.5 ions that are readily ascribed to [Mo₆Cl₁₃Br]²⁻ ions. The observation of these Cl⁻ and Br⁻-containing ions suggests that a ligand-exchange reaction occurred between [Mo₆Cl₁₃]⁻ and Br⁻-containing ions. At this stage of the study, we can of course not define whether this reaction is occurring in solution or in the gaseous ion beam.

Fig. 3 (a) Negative ESI-MS full-scan MS spectrum of a solution of Cs₂[{Mo₆Clⁱ₈}Cl^a₆] and cesium bromide. A signal magnification of 150 times was done in the m/z 970–1070 range. The major signals at m/z 930.2 and 718.4 correspond to the [Br₅Cs₄]⁻ and [Br₄Cs₃]⁻, respectively. (b) Experimental (top) and theoretical (bottom) isotopic patterns of the [Mo₆Cl₁₃Br]²⁻ ion (m/z 545–575).

Furthermore, taking advantage of the fact that [Cs₄KBr₆]⁻ ions and [Mo₆Cl₁₃]⁻ are characterized by proximal m/z values, with their isotopic patterns overlapping in the m/z 970–1070 range, these anions may be concomitantly mass-selected in the course of CID experiments by enlarging the quadrupole mass selection window, i.e. from m/z 1010 to m/z 1070. The resulting mass spectrum that is represented in Fig. 4 shows not only the generation of the [{Mo₆Clⁱ₈}Cl^a₆]²⁻ ions (i.e. chlorine addition on the mass-selected ions) as observed previously when selecting only [Mo₆Cl₁₃]⁻ ions, but also the formation of the halogen-mixed [Mo₆Cl₁₃Br]²⁻ ions. We propose to attribute the formation of the [{Mo₆Clⁱ₈}Cl^a₆]²⁻ to the reactive self-interaction under collision activation of the [Mo₆Cl₁₃]⁻ selected ions (Scheme 1). Similarly, the [Mo₆Cl₁₃Br]²⁻ ions are likely to be issued from the reaction in the T-Wave™ cell of the mass spectrometer between [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ and [Cs₄KBr₆]⁻ ions, as described in Scheme 1. Besides the [Mo₆Cl₁₃Br]²⁻ isotopic pattern centered on the m/z 558.5 value, other signals detected support this assumption (Fig. 4). If the ions at m/z 836.3 ([Cs₃KBr₅]⁻, Fig. S2, ESI†), m/z 718.4 ([Cs₃Br₄]⁻, Fig. S3, ESI†) and m/z 624.4 ([Cs₂KBr₄]⁻, Fig. S4, ESI†) can be attributed to the dissociation of the selected [Cs₄KBr₆]⁻ ion, the ions at m/z 674.4 ([Cs₃Br₃Cl]⁻, Fig. S5, ESI†), m/z 628.5 ([Cs₃Br₂Cl₂]⁻, Fig. S6, ESI†) and m/z 584.4 ([Mo₆Cl₁₂Br₂]²⁻, Fig. S7, ESI†) are ionic species produced by Cl⁻ and Br⁻ exchanges between the concomitantly mass-selected two precursor ions. It is unlikely that halide exchanges would arise by an initial dissociation step of both [Mo₆Cl₁₃]⁻ and [Cs₄KBr₆]⁻ ions followed by Br⁻ capture by the former and the Cl⁻ capture by the fragments of the latter. [Cs₃Br₂Cl₂]⁻ (m/z 628.5) and [Mo₆Cl₁₂Br₂]²⁻ (m/z 584.4) ions could be observed if such processes occur. Finally, [CsMo₆Cl₁₄]⁻ (Fig. S8, ESI†) whose isotopic pattern is observed around the m/z 1209.3 value (Fig. 4), involves the transfer of a Cs⁺ cation and a Cl⁻ anion that can only be due to an interaction between [Mo₆Cl₁₃]⁻ and [Cs₄KBr₆]⁻ ions. It has to be noted that the ions produced by the halide exchanges during the collision activation are not detected anymore upon increasing the trap collision voltage to 5 V. This experiment unambiguously demonstrates that the reactions occur in the gaseous ion beam.

Fig. 4 Negative ESI-MS-MS spectrum of the [Mo₆Cl₁₃⁻ and [Cs₄KBr₆]⁻ ions selected around m/z 1036 (see ESI† for more details).

Gas phase nucleophilic addition reactions between anion clusters such as [B₁₂L₁₁]⁻ (L = F, Cl, I or CN) and inert molecules such as argon, krypton, neon and N₂ have already been described in several MS experiments attesting to the electrophilic character of the anion clusters.^8–10 This behavior is attributed by the authors to a strong electronic density depletion at the active site, qualifying these ions as “dipole-discriminating electrophilic anions” on the basis of computational results.¹⁰ More importantly, in the same family of compounds, [B₁₂I₉]⁻ anions have been shown to dimerize within an ion trap collision cell giving rise to [B₂₄I₁₈]²⁻ dianions.^19,20 These results unambiguously prove the possibility of interactions between anions in MS experiments.

In the present study, it is a bond breaking that is observed. Indeed, CID-MS-MS experiments show that Br⁻ is transferred from [Cs₄KBr₆]⁻ to [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ to yield [{Mo₆Clⁱ₈}Cl^a₅Br^a]²⁻. The reactivities of [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ in tandem-in-space MS experiments and the dimerization of [B₁₂I₉]⁻ anions in ESI-MS are counter-intuitive at first sight since the ion density is low and anions should repel each other considering that they are negatively-charged. In the present CID-MS-MS experiments, since the anions are moving in the same direction should additionally prevent interactions. A few studies of ion trajectories through high pressure quadrupole cell modelling are available.^21,22 Lock et al.²² established that only the quadrupole collision cell becomes thermalized in such experiments considering that ions are crossing a relatively high pressure radio frequency cell. Ions could thus exit the device via the repulsive effects between them while streaming into the entrance of the cell.

The clue comes from the study of the electrostatic and electronic features of the [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ ions. As shown in the computational section, there is a heterogeneous distribution of positive and negative charges at the surface of the monocharged [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ anions. The electronic density depletion observed around the lacunary molybdenum atom is similar to what is reported in the [B₁₂L₁₁]⁻ series (see electrostatic potential mapping in Fig. 1). This allows concluding that no coulombic repulsion occurs if the interaction with a negatively-charged ion happens in the vicinity of the positively-charged parts of the [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ cluster ions. Quantum chemical calculations reveal that the lacunary molybdenum site of the [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ mono-anion is moreover highly sensitive to nucleophilic attacks (Fig. 1).

The detection of [{Mo₆Clⁱ₈}Cl^a₅Br^a]²⁻ species when isolating the [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ and [Cs₄KBr₆]⁻ species concomitantly reveals that the energy acquired by the ions is sufficient to subtract Br⁻ from [Cs₄KBr₆]⁻. This leaves a neutral [Cs₄KBr₅] entity, which cannot be detected by MS. Note that, due to the lack of structural data for the [Cs₄KBr₆]⁻ ions, the energy requirement of the transfer reaction is not yet established, making the ion kinetic energy dependency on the Br⁻ transfer reaction efficiency still to be evaluated. Importantly, this result reveals that there is a confinement effect in the T-Wave™ cell or the travelling wave ion guide (TWIG) trap that is sufficient to allow ion interaction as already pointed out by Lock and coworkers in the case of high pressure cells.^21,22

The same process is involved in the formation of [{Mo₆Clⁱ₈}Cl^a₆]²⁻, which consists in a heterolytic Mo–Cl bond breaking in [{Mo₆Clⁱ₈}Cl^a₅□^a]− by interaction with a second [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ ion. In that case, the required energy to subtract Cl⁻ from [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ is estimated to be 263 kJ mol⁻¹ by quantum chemical calculations. This is of the order of magnitude of what can be evaluated in terms of total internal energy imparted in the precursor ion during a collision process such as described in the present experiments (see ESI† for details). The computational study reveals that [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ is indeed susceptible to electrophilic attack on the Cl ligand in the trans position to its electrophilic site (Fig. 1). The alignment of the dipole moments of [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ ions (6.52 Debye) with the electric field induces that the electrophilic sites and nucleophilic sites of two consecutive [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ ions are positioned to avoid coulomb repulsion if interacting (Fig. 1).

Such superelectrophilic behaviors are reported for cations;²³ the only anions known to date are the [B₁₂L₁₁]⁻ series which activate inert gas. Dimerization of [B₁₂I₁₁]⁻ is the only anion–anion interaction known in the gas phase.^8–10 Herein, we demonstrate that such electrophilic behavior exists for [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ and that it leads to an iono-covalent bond creation with another anion. The ideal molecular orientation induced by the alignment of the dipole moment and the applied electric field in the CID allows the electron-donating anions to interact without coulombic repulsion. Interestingly, [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ ions have their nucleophile site in the trans position to its electrophilic site. This makes the [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ ion a unique example in the inorganic chemistry of Janus anions exhibiting both nucleophilic and electrophilic character. This opens up a wide range of fundamental studies toward the activation of cluster units in the gas phase and further multi-reactivity. Functionalization of a surface by [{Mo₆Clⁱ₈}Cl^a₅□^a]⁻ ions in a gas phase process becomes possible. It can provide for instance sensitive layers able to interact with volatile organic compounds for the design of antenna sensors such as some of us have recently patented.²⁴

This work was granted access by the French HPC agency GENCI (TGCC/CINES/IDRIS – A0100800649/AD010800649R1). The authors warmly thank Rémi Marchal for technical support.

Conflicts of interest

There are no conflicts to declare.

Notes and references

1. N. T. K. Nguyen, C. Lebastard, M. Wilmet, N. Dumait, A. Renaud, S. Cordier, N. Ohashi, T. Uchikoshi and F. Grasset, Sci. Technol. Adv. Mate., 2022, 23, 547–578.
2. K. Kirakci, M. A. Shestopalov and K. Lang, Coord. Chem. Rev., 2023, 481, 215048.
3. A. Pinkard, A. M. Champsaur and X. Roy, Acc. Chem. Res., 2018, 51, 919–929.
4. P. Jena and Q. Sun, Chem. Rev., 2018, 51, 5755–5870.
5. A. W. Castleman and S. N. Khanna, J. Phys. Chem. C, 2009, 113, 2664–2675.
6. P. Su, Z. Warneke, D. Volke, M. F. Espenship, H. Hu, S. Kawa, K. Kirakci, R. Hoffmann, J. Laskin, C. Wiebeler and J. Warneke, J. Am. Soc. Mass Spectrom., 2023, 34, 161–170.
7. V. K. Kochnev, V. V. Avdeeva, E. A. Malinina and N. T. Kuznetsov, Russ. J. Inorg. Chem., 2014, 59, 1268–1275.
8. M. Rohdenburg, M. Mayer, M. Grellmann, C. Jenne, T. Borrmann, F. Kleemiss, V. A. Azov, K. R. Asmis, S. Grabowsky and J. Warneke, Angew. Chem., Int. Ed., 2017, 56, 7980–7985.
9. M. Mayer, V. van Lessen, M. Rohdenburg, G. L. Hou, Z. Yang, R. M. Exner, E. Apra, V. A. Azov, S. Grabowsky, S. S. Xantheas, K. R. Asmis, X. B. Wang, C. Jenne and J. Warneke, Proc. Natl. Acad. Sci. U. S. A., 2019, 116, 8167–8172.
10. M. Rohdenburg, V. A. Azov and J. Warneke, Front. Chem., 2020, 8, 580295.
11. M. M. Zhong, H. Fang, Deepika and P. Jena, Phys. Chem. Chem. Phys., 2021, 23, 21496–21500.
12. K. Morokuma, J. Chem. Phys., 1971, 55, 1236.
13. K. Kitaura and K. Morokuma, Int. J. Quantum Chem., 1976, 10, 325–340.
14. T. Ziegler and A. Rauk, Theor. Chim. Acta, 1977, 46, 1–10.
15. T. Ziegler and A. Rauk, Inorg. Chem., 1979, 18, 1755–1759.
16. T. Ziegler and A. Rauk, Inorg. Chem., 1979, 18, 1558–1565.
17. M. Raupach and R. Tonner, J. Chem. Phys., 2015, 142(1–14), 194105.
18. R. G. Parr and W. T. Yang, J. Am. Chem. Soc., 1984, 106, 4049–4050.
19. J. Warneke, T. Dulcks, C. Knapp and D. Gabel, Phys. Chem. Chem. Phys., 2011, 13, 5712–5721.
20. T. S. Zhao, J. Zhou, Q. Wang and P. Jena, J. Phys. Chem. Lett., 2016, 7, 2689–2695.
21. A. G. Harrison, Rapid Commun. Mass Spectrom., 1999, 13, 1663–1670.
22. C. M. Lock and E. W. Dyer, Rapid Commun. Mass Spectrom., 1999, 13, 422–431.
23. D. A. Klumpp and M. V. Anokhin, Molecules, 2020, 25(1–25), 3281.
24. A. Denis, D. Rondeau, M. Himdi, X. Castel and S. Cordier, Antennas having ion cluster type sensors, WO2021/175796A1 (PCT/EP2021/055080), 2021.

---

Footnote

† Electronic supplementary information (ESI) available: Syntheses, experimental and computational details, MS spectra. See DOI: https://doi.org/10.1039/d3cc01187a

---