Macropolyhedral syn-B18H22, the “Forgotten” Isomer

The chemistry and physics of macropolyhedral B18H22 clusters have attracted significant attention due to the interesting photophysical properties of anti-B18H22 (blue emission, laser properties) and related potential applications. We have focused our attention on the “forgotten” syn-B18H22 isomer, which has received very little attention since its discovery compared to its anti-B18H22 isomer, presumably because numerous studies have reported this isomer as nonluminescent. In our study, we show that in crystalline form, syn-B18H22 exhibits blue fluorescence and becomes phosphorescent when substituted at various positions on the cluster, associated with peculiar microstructural-dependent effects. This work is a combined theoretical and experimental investigation that includes the synthesis, separation, structural characterization, and first elucidation of the photophysical properties of three different monothiol-substituted cluster isomers, [1-HS-syn-B18H21] 1, [3-HS-syn-B18H21] 3, and [4-HS-syn-B18H21] 4, of which isomers 1 and 4 have been proved to exist in two different polymorphic forms. All of these newly substituted macropolyhedral cluster derivatives (1, 3, and 4) have been fully characterized by NMR spectroscopy, mass spectrometry, single-crystal X-ray diffraction, IR spectroscopy, and luminescence spectroscopy. This study also presents the first report on the mechanochromic shift in the luminescence of a borane cluster and generally enriches the area of rather rare boron-based luminescent materials. In addition, we present the first results proving that they are useful constituents of carbon-free self-assembled monolayers.


Contents NMR
.Experimental Stack plot of 11 B{ 1 H} decoupled spectra of syn-B 18 H 22 , 1-HS-syn-B 18 H 21 , 3-HS-syn-B 18 H 21 , and 4-HS-syn-B 18 H 21 .                         Figure S34. Overview X-ray photoelectron spectrum (A) as well as high-resolution C 1s (B) and O 1s (C) spectra of the formed [4-HS-syn-B 18 H 21 ] 4 SAM on Ag. The data clearly show only the presence of boron, sulfur and silver signals in the sample. No carbon or oxygen signals are detected demonstrating a high quality of the formed SAM.      . Powder X-ray diffraction pattern (PXRD) of anti-B 18 H 22 (obtained at room temperature using a PANalytical Empyrean diffractometer equipped with a conventional Cu X-ray tube and PIXcel detector). The data were processed using the HighScore Plus program from PANalytical, and the obtained diffraction pattern fits perfectly to the single-crystal simulated one. Figure S41. ORTEP structure of syn-B 18 H 22 isomer having 50% thermal ellipsoid probability. Figure S42. ORTEP structure of 1-HS-syn-B 18 H 21 PM1a isomer having 50% thermal ellipsoid probability. Figure S43. ORTEP structure of 1-HS-syn-B 18 H 21 PM1b isomer having 50% thermal ellipsoid probability. Figure S44. ORTEP structure of 3-HS-syn-B 18 H 21 isomer having 50% thermal ellipsoid probability. Figure S45. ORTEP structure of 4-HS-syn-B 18 H 21 PM4a isomer having 50% thermal ellipsoid probability. Figure S46. ORTEP structure of 4-HS-syn-B 18 H 21 PM4b isomer having 50% thermal ellipsoid probability.  Table S8. Crystallographic collection and refinement data for 1-HS-syn-B 18 H 21 (PM1a). Table S9. Crystallographic collection and refinement data for 1-HS-syn-B 18 H 21 (PM1b). Table S10. Crystallographic collection and refinement data for 3-HS-syn-B 18 H 21 . Table S11. Crystallographic collection and refinement data for 4-HS-syn-B 18 H 21 (PM 4a). Table S12. Crystallographic collection and refinement data for PM4a measured at room temperature. Table S13. Crystallographic collection and refinement data for 4-HS-syn-B 18 H 21 (PM4b).   Figure S48: Simulated spectra (A, D), excited states (B, E) and excited states in the relaxed geometry of S1 (C, F) of anti (A-C) and syn (D-F) B 18 H 22 ; in the excited states diagrams, singlets are depicted yellow, triplets orange, and the length of the yellow lines on the left side, with scale labels in nm, represents the relative oscillator strengths of the singlet excited states. Figure S49: Simulated spectra (A), excited states (B) and excited states in the relaxed geometry of S1 (C) of 1-HS-syn-B 18 H 21. Figure S50: Simulated spectra (A), excited states (B) and excited states in the relaxed geometry of S1 (C) of 2-HS-syn-B 18 H 21. Figure S51: Simulated spectra (A), excited states (B) and excited states in the relaxed geometry of S1 (C) of 3-HS-syn-B 18 H 21. Figure S52: Simulated spectra (A), excited states (B) and excited states in the relaxed geometry of S1 (C) of 4-HS-syn-B 18 H 21.   : Natural Transition Orbitals (bottom hole, top electron) for the absorption of syn-B 18 H 22 to the first excited singlet in the ground state geometry (left) and its luminescence from the relaxed geometry of S1 (right). Table S15: Excitation energies (E exc ) and the associated wavelengths (λ) of the first ((S1) and second (S2) excited states in the ground state geometry, and that of the first singlet excited state in its optimised geometry (S1/S1)
The 11 B COSY NMR spectrum ( Figure S5) allows to deduce information on the connectivity of the boron atoms represented by individual signals. The signal a has cross peaks with e, f, h, i, and j; b with i, and possibly a weak one with h; c with e, g, and j; d only with j; e with a, c, g, i, and j; f with a, h, and j; g with c, e, and i; h with a, f, i, and possibly b; i with a, b, e, g, and h; and j with a, c, d, e, and f. So a will be a neighbour of e, and they will have two common neighbours, i and j, which suggests the arrangement of two triangles a-e-i and a-e-j with a-e for a common edge; a and i have h for another common neighbour besides e, a and j form another triangle with f, e and i with g, and e and j with c, while h and f also interact with one another, and so do c and g; we can connect d on the basis of the COSY only to j, and b to i with a possible but uncertain completion of a triangle with h ( Figure  S6).
The 11 B-1 H HMQC spectrum ( Figure S8) allows assignment of protons to borons. To identify the correct position of our connection diagram in the schematic structure of syn-B 18 H 22 ( Figure 1B), we will need, especially for the bridging hydrogens, the information which pairs of boron atoms are connected by them. Among our 11 B signals, a has a cross peak with 1 H at 3.8 ppm, b at 1.3 ppm, and we know for two reasons that this is a bridging hydrogen: b is a singlet, so it has no terminal hydrogen, and the hydrogen at 1.3 has another cross peak with h; at the same time, h has another cross peak with 1 H, its terminal hydrogen, at 2.6 ppm. For c, there is a cross peak with -0.9 ppm with another bridging hydrogen that interacts with d as well, and a weak cross peak with its terminal 1 H at 3.8 ppm. The cross peak with the terminal 1 H of d is at 3.4 ppm, and there is one more cross peak of d with one more bridging hydrogen at -2.5 ppm; that bridging hydrogen interacts with f. That boron has the cross peak with its terminal hydrogen at 3.1 ppm, while e has it at 3.6 ppm, i at -0.2 ppm, j at 0.6 ppm, and g, the second singlet in 11 B spectrum, has none. The important part for 11 B NMR assignment are the three bridging hydrogens that connect b with h, c with d, and d with f: when we add these hydrogen bridges to our connection diagram, we can finally fit it on the syn-B 18 H 22 structure unequivocally ( Figure S7).

1-HS-syn-B 18 H 21 11 B NMR assignment
In the 11 B NMR spectra of monosubstituted syn-B 18 H 22 derivatives we can expect three singlets: the peaks of B(5) and B(6), and that of the substituted atom. In the spectra of 1-HS-syn-B 18 H 21 ( Figure S10) we see clearly two singlets at 11.0 ppm and -3.8 ppm, altogether 16 major peaks, and three smaller peaks of a major impurity (10-20 molar %) that correspond to three peaks (of intensity 2) of the parent syn-B 18 H 22 . The peaks at 8.3 ppm and -28.1 ppm have integrated intensity two times larger than the other 14, and that at 8.3 ppm contains the third singlet, coinciding with a doublet. Comparing the spectra of the derivative with those of the parent borane, we can tentatively assign the pair of peaks in the lowest field, at 16.7 ppm and 15.0 ppm, to the boron atoms 3 and 3', that in the highest field, at -37.2 and -40.6 ppm, to the borons 4 and 4', the peaks coinciding at -28.0 ppm to 2 and 2', and the pair of peaks at -14.3 and -15.6 ppm to 7 and 7'. If we did not know from the X-ray structure that the compound is 1-HS-syn-B 18 H 21 , we could have deduced it from our chemical expectation that the substitution would occur in one of the four positions 1, 2, 3, and 4, and the fact that all the peaks in the areas around those of the borons 3, 4, and 2 of the parent borane are doublets here. We can expect that the singlet at the highest field, that at -3.8 ppm, will belong to the boron 5 (-4.1 ppm in the parent), while choosing between the assignment of 6 (9.7 ppm in the parent) and 1 (1.7 in the parent) to the singlets at 11.0 ppm and 8.3 ppm is trickier, as the 11 B COSY shows.
In the 11 B-11 B COSY ( Figure S12), we may not see all the cross peaks we could expect according to what we have seen in the COSY of the parent, but after the analysis of the spectrum we were able to fine tune its transformation so that we can find at least tentative traces of most, keeping still the occurrence of noise and artefacts at an acceptable level ( Figure S13). For 5, we expect cross peaks with 1, 1', 2, 2', and 10, 10', for 6 with 2, and 2' only, and for 1 with 2, 3, 4, 5, and 10, and no one with any atom of the unsubstituted subunit. The singlet at -3.8 ppm has a strong cross peak with that at 11.0 ppm, one more strong one at 2.0 ppm, a weaker one with the peak of 2, 2', and a weak one at 8.6 ppm, with one of the peaks coinciding at 8.3 ppm in the 1D spectrum. That is in accord with the assignment of -3.8 ppm to the boron 5, and points to the singlet at 11.0 ppm belonging to the boron 1, not 6. This is confirmed by the singlet at 11.0 ppm having multiple cross peaks besides that with 5: also with 16.7 ppm (so it is 3, and 15.0 ppm will be 3'), -37.2 ppm (4, and -40.6 ppm will be 4'), -28.0 ppm (2, coinciding with 2'), and with one of the peaks coinciding at 8.3 ppm -only 10 remains to be assigned, and this position corresponds to the cross peak of 5 at 8.6 ppm, the exact position of which is possibly affected there by the neighbouring strong cross peak of 5 with 1. There is another cross peak of the coinciding pair of signals at 8.3 ppm with 2, 2' around [8.3 ppm, -28.0 ppm], and even one uncertain weak one with 4 at [8.3 ppm, -37.2 ppm], so as those signals coinciding at 8.3 ppm should be those of the borons 6 and 10, for 6 we expect the only cross peak with 2, 2', and for 10 those with 1, 4, and 5, we should assign the chemical shift of 8.3 ppm to both borons 6 and 10, their coincidence exact within the error of our measurement.
As the boron 3 in the parent borane has cross peaks with 1, 2, 4, 7, and 8, we can, after the assignment of 11.0 ppm to 1, and consequently 16.7 ppm to 3 and -37.2 ppm to 4 follow the cross peaks of 3 (16.7 ppm) and 3' (15.0 ppm), confirm the assignment of -37.2 ppm to 4 and -40.6 ppm to 4', assign -28.0 ppm to 2 and -28.4 ppm to 2', -14.3 ppm to 7 and -16.6 to 7', and 2.0 ppm to 1', and end up assigning the weak cross peaks at 0.0 ppm to 3 with 8, and at -2.0 ppm to 3' with 8'; we can verify this assignment by looking up other weak cross peaks of 8 with 7, 8' with 7', and 8' with 4', that of 8 with 4 possibly present as well, but hidden by the vicinity of a much stronger cross peak. The assignment of 2.0 ppm to 1' is confirmed by its cross peaks with 5, 2', and 4', the last unassigned of them at 9.7 ppm thus belonging necessarily to 10'. Boron 4 in the parent borane interacts with 1, 3, 8, 9, and 10, here we see cross peaks of 4 with 1 and 3, have already discussed the weak one with 10 and apparent absence of that with 8, so the last one, strong enough, at 1.1 ppm remains for 9; analogically we have already assigned the strong cross peaks of 4' with 1' and 3', as well as the weak ones with 10' and 8', so the last unassigned strong cross peak 3.3 ppm must correspond to 9'.
We The cross peaks we see in the 11 B-1 H HMQC NMR spectrum ( Figure S14) do not allow us to assign 1 H NMR shifts to all the hydrogen atoms in the molecule. Looking up the cross peaks of the boron atoms from 3 (16.7 ppm) to 4' (-40.6 ppm), we find the 1 H chemical shifts 4.2 ppm for 3, and 3.9 ppm for 3'. The substituted boron 1 has no terminal hydrogen, but 10' does, and we see only its cross peak with the bridging hydrogen 10'-9' at -1.0 ppm anyway. The case of 10 and 6 is identical, μH(10, 9) has the chemical shift -0.4 ppm, and μH(6, 7) 1.7 ppm; even the expected cross peak of B(6) with μH(6, 7') is absent. The boron 9' has, besides its cross peak with μH(10', 9') that with μH(9', 8') at -2.5 ppm, but again none with a terminal hydrogen, only for 1' we see the cross peak with H(1') at 3.7 ppm. μH(9, 8) has chemical shift -2.5 ppm but only B(8') has also a visible cross peak with its terminal hydrogen at 3.1 ppm. For the boron 5 we do not expect contact with any hydrogen but at the borons 7 and 7' we again have expected cross peaks missing, 7 only having on with μH(6, 7), and 7' in contrast with H(7') at 2.6 ppm. According to the only cross peak of the coinciding signals of B(2) and B(2'), H(2) and H(2') probably have the same chemical shift -0.2 ppm, and the cross peaks of the borons 4 and 4' show the chemical shifts of H(4) as 0.7 ppm and of H(4') as 0.6 ppm. At least the five bridging hydrogens identified of the expected six confirm our assignment of the 11 B spectrum.

3-HS-syn-B 18 H 21 11 B NMR assignment
In the comparison of 1 H decoupled and simple (coupled) 11 B NMR spectra of 3-HS-syn-B 18 H 21 ( Figure S16) we see immediately three singlets, that we can on the basis of the parent borane spectrum assign to the substituted position 3 (21.4 ppm) and the two borons common to both subunits, that have no terminal hydrogens, 6 (6.6 ppm) and 5 (-2.2 ppm). The decoupled spectrum contains 17 peaks, in one of them the singlet of B(5) coincides with a doublet of another boron atoms, and all the signals of the other 16 boron atoms are individually resolved. Still by the analogy with the spectra of the parent we can assign 15.0 ppm to 3' and tell that the three pairs in the upfield part of the spectrum belong -15.4 ppm and -16.8 ppm to 7 and 7', -28.6 ppm and -29.4 ppm to 2 and 2', and -38.5 ppm and -39.4 ppm to 4 and 4'.
Again, we verify and complete the assignment by interpreting the 11 B-11 B COSY NMR spectrum ( Figure S18). Cross peaks of 3 and 3' allow us to assign -38.5 ppm to 4 and -39.4 ppm to 4', -28.6 ppm to 2 and -29.4 ppm to 2', -29.4 ppm to 7 and -28.6 ppm to 7', and 4.4 ppm to 1 and 2.7 ppm to 1', and -3.2 ppm to 8, and -2.0 ppm to 8', in the 1D spectrum coinciding with 5 at -2.2 ppm. As we expect for 5 cross peaks with 1, 1', 2, 2', 10, and 10', for 8' with 3', 4', and 7', and for 1' with 2', 3', 4', 5, and 10', our assignment is verified by the cross peaks at [-2.2 ppm, 4.4 ppm] and [-2.2 ppm, 2.6 ppm] that can only belong to 5-1 and 5-1' because 8 and 8' cannot have cross peaks with 5, nor with 1'. We do not see the cross peaks of 5 with 10 and 10' but 1 has a weak cross peak with 10 at 10.3 ppm, and 1' has similar with 10' at 9.7 ppm. Even though we do not see all the expected cross peaks for 4 and 4', there are strong enough cross peaks of 4 with 9 at 0.8 ppm and 4' with 9' at 9.7 ppm. The assignment of the 11 B spectrum is thus complete, and other visible cross peaks like those of 5 with 2 and 2' of 6 with 2 and 2' or those of 1 with 2 and 1' with 2' corroborate it.
An interesting feature in the 11 B-1 H HMQC spectrum of 3-HS-syn-B 18 H 21 ( Figure S19) is the cross peak of the substituted B(3) at 1.4 ppm -with the thiol group hydrogen, in the 1 H NMR spectrum decoupled by 11 B represented by a very sharp peak, and without decoupling only slightly broadened ( Figure S20). 3' has a cross peak with its terminal hydrogen at 3.8 ppm. For 10 and 10' we have detected cross peaks only with their bridging hydrogens with B(9) and B(9'); the chemical shifts of those hydrogens are -0.9 ppm, identical within our experimental error, even though μH(9, 10) is visibly upfield of μH(9', 10'). B(6) has strong cross peaks with its bridging hydrogens, μH(6, 7) at 1.9 ppm, downfield from μH(6, 7') at 1.1 ppm. According to the cross peaks of 1 and 1', H(1) has chemical shift 4.0 ppm, and H(1') 3.6 ppm. For borons 9 and 9' we see only their cross peaks with their bridging hydrogens, besides μH(9, 10) and μH(9', 10'), there are μH(8, 9) at -2.1 ppm and μH(8', 9') at -2.5 ppm. As B (5) is not connected to any hydrogen atom, it does not interfere with the assignment of the chemical shift 3.5 ppm to H(8´), while H(8) has 3.1 ppm, according to a weak cross peak. The borons 7 and 7' have, besides their cross peaks with μH(6, 7) and μH(6, 7') those with their terminal hydrogens too, even though that of B (7) with H(7) at 2.6 ppm is weaker than that of B(7') with H(7') at 2.7 ppm, just like that of B (8)

4-HS-syn-B 18 H 21 11 B NMR assignment
4-HS-syn-B 18 H 21 has many 11 B NMR signals ( Figure S22) very close together, especially the group 1, 1', 9, 9'; 8 and 8' almost coincide, and 7, 7' do not differ at all. Nevertheless, a strong line narrowing transformation (Gaussian window on the FID before the Fourier transform 42 Hz wide and shifted 13 ms for the 1 H decoupled spectrum and 50 Hz shifted 10 ms for the coupled one) allows to differentiate all the peaks ( Figure S23). Interestingly, there is at -25.8 ppm, in the typical area of 2, 2' signals, a singlet, so likely a signal of a substituted B(2), which suggests that even the 2-HS isomer that we have not obtained in pure form may have been form at least as a very minor product, and present as an impurity.
Singlets at 8.1 ppm, -3.4 ppm, and -29.2 ppm can be assigned to B(6), B(5), and B(4) respectively just on the comparison with the spectra of the parent and the other isomers: 8.1 ppm is upfield of B(6) of the parent, slightly more than in 1-HS and less than in 3-HS, -3.4 ppm is downfield of B(5) in the parent, again more than in 1-HS but less than in 3-HS, and -29.2 ppm is a singlet downfield of B(4) peaks at the upfield end of syn-B 18 H 22 derivatives spectra, shifted from the parent towards the B(2) region, but must belong to the substituted B(4), even if we did not know from the X-ray that this is the 4-HS isomer, because there are two doublets close to one another at -27.7 ppm and -28.1 ppm, evidently B(2) and B(2') (we only have to find out which is which), and only one doublet of B(4') is left at -39.8 ppm, downfield of 4, 4' of the parent but less than B(4') of 3-HS.
In 11 B-11 B COSY, B(4) of syn-B 18 H 22 ( Figure S5) has cross peaks with 1, 2, 3, 8, 10, and 9. In the COSY of 4-HS-syn-B 18 H 21 ( Figure S25), no 4'-2' or 4-2 cross peak is seen. There are four cross peaks of 4 (or 5 if there are two at 3.0 and 2.6 ppm fused together), and four of 4', so besides those with 2 and 2', one more must be missing, at least for 4', perhaps that with 8', based on our experience with 1-HS and 3-HS. Assignment of 14.9 ppm to 3 and 13.8 ppm to 3' is easy, both downfield of the parent, 3 about as much as 3' in 1-HS, 3' less. Similarly, 10.3 ppm will be B(10), downfield of the parent about as much as in 3-HS, and 9.1 ppm B(10'), slightly upfield of the parent, but less then B(10) in 1-HS. 4' has a strong cross peak with 3.5 ppm and weaker with 2.0 ppm, 4 has a strong one with 2.6 ppm, and possibly a weaker one with 3.0 ppm, fused with the stronger one. We can assign them, along with 2 and 2', based on the cross peaks of 2 and 2', as this is the area of borons 1 and 9, and in the parent borane, 9 has a cross peak only with 4 (and can hardly have one with 2, not even hypothetically), while 1 has cross peaks with both 4 and 2 (and 3, 5, and 10). The signal at -28.1 ppm has a cross peak with 2.0 ppm, which has a cross peak with 4', so 2.0 ppm is 1', and -28.1 ppm is 2'. It leaves 2 to assign to -27.7 ppm, and its cross peak with 3.0 ppm, that may, with some uncertainty, have one with 4 as well, make 3.0 ppm B(1). Subsequently 3.5 ppm must belong to B(9'), and 2.6 ppm to B(9). The remaining cross peak of B(4) at -1.2 ppm allows us to assign -1.2 ppm to B(8), and consequently -1.6 ppm, even without cross peaks detected, to B(8'). The cross peaks of 1 with 3 and 10, and of 1' with 3' and 10', corroborate this assignment, as do the cross peaks of 3 with 2 and 3' with 2', borons 5 and 6 having cross peaks with both 2 and 2'. The peak at -16.1 ppm that belongs to borons 7 and 7', has cross peaks only with 3 and 3'; those allow to differentiate a bit, albeit around the level of experimental error, between B(7) and B(7'), and to assign -16.1 ppm to 7', and -16.2 ppm to 7. We thus have the signals in the 11 B ( 1 H decoupled) NMR spectrum of 4-HS-syn-B 18 H 21 assigned as: 3 (14.9 ppm) > 3' (13.8 ppm) > 10 (10.3 ppm) > 10' (9.1 ppm) > 6 (8.1 ppm) > 9' (3.5 ppm) > 1 (3.0 ppm) > 9 (2.6 ppm) > 1' (2.0 ppm) > 8 (-1.              (2) Powder X-ray diffraction and Rietveld refinement XRPD was investigated on a SmartLab Rigaku powder diffractometer with a rotating Cu anode at room temperature using a HyPix-3000 detector. It was confirmed that no phase of HS-syn-B 18 H 21 was missed. The data were processed using the HighScore Plus program from PANalytical. The results of the phase analysis are in following Figures S36 to S40.     . Powder X-ray diffraction pattern (PXRD) of anti-B 18 H 22 (obtained at room temperature using a PANalytical Empyrean diffractometer equipped with a conventional Cu X-ray tube and PIXcel detector). The data were processed using the HighScore Plus program from PANalytical, and the obtained diffraction pattern fits perfectly to the single-crystal simulated one (CCDC number 602765).

Single crystal X-ray analysis
The X-ray diffraction data of samples were collected with Rigaku OD Supernova using Atlas S2 CCD detector and mirror collimated Cu-Kα (λ = 1.54184 Å) from micro-focused sealed X-ray tube. The samples were cooled to 95 K during the measurement. Integration of the CCD images, absorption correction and scaling were done by program CrysAlisPro 1.171.41.123a (Rigaku Oxford Diffraction, 2022). Crystal structures were solved by charge flipping with program SUPERFLIP and refined with the Jana2020 program package by full-matrix least-squares technique on F 2 . All hydrogen atoms were visible in difference Fourier maps and positions of all of them were refined freely. The isotropic atomic displacement parameters of hydrogen atoms were evaluated as 1.2Ueq of the parent atom. ORTEP diagrams ( Figure S41 to S46) were made in Diamond 4.5.1. Obtained crystallographic data are in Tables S6 to S13 Figure S41. ORTEP structure of syn-B 18 H 22 isomer having 50% thermal ellipsoid probability.      (1) The refinement program Jana2020 does not refine the weighting scheme. Therefore, GOF is usually fairly above one, especially for strongly diffracting crystals.

TD-DFT
The results of the Time Dependent Density Functional Theory (TD-DFT) calculations are summarised in the Table S. Simulated spectra and calculated excited states energies of anti-and syn-B 18 H 22 are compared in the Figure S48, those of 1-, 2-, 3-, and 4-HS-syn-B 18 H 21 follow in Figure S49 to Figure S52.  Table S15: Excitation energies (E exc ) and the associated wavelengths (λ) of the first ((S1) and second (S2) excited states in the ground state geometry, and that of the first singlet excited state in its optimised geometry (S1/S1) E exc (S1)/eV λ(S1)/nm E exc (S2)/eV λ(S2)/nm E exc (S1/S1)/eV λ(S1/S1)/nm Upon optimisation of the first excited singlet geometry on the TD-DFT level of theory, the calculated energy of the first excited singlet of syn-B 18 H 22 above the ground state lowers from 4.56 eV only to 3.68 eV, the S1 thus remains in its relaxed geometry even a bit higher above the ground state than that of anti-B 18 H 22 . The differences between the optimized geometry of the ground state, and that of the first excited singlet, are for both anti-B 18 H 22 ( Figure S53) and syn-B 18 H 22 ( Figure S54) almost imperceptible. There is no indication in our TD-DFT results pointing to the existence of an easily accessible conical intersection of the first excited singlet with the ground state, claimed as responsible for the observed absence of luminescence in the solution of syn-B 18 H 22 at the CASPT2/CASSCF level of theory i . For a correct description of the excited states of syn-B 18 H 22 , extensive theoretical studies at higher levels of theory are probably necessary. Also, the calculated luminescence wavelengths λ(S1/S1) of syn-B 18 H 22 and all its thiol derivatives investigated are too short compared to those measured experimentally in the solid state. Similarly, the longest calculated absorption wavelength, the weak line at 337 nm of 4-HS-syn-B 18 H 21 (Table S15, Figure S52), lies still in the ultraviolet range, although the species forms intensively pink crystals. These observations clearly suggest that the spectroscopic properties of these species in solid state are predominantly determined by their supramolecular crystalline structure, and their further theoretical analysis is beyond the scope of the basic DFT calculations of isolated molecules.