10.1002/chem.201901450
Chemistry - A European Journal
COMMUNICATION
idiosyncrasies of the alumina mediated HF-elimination. First of
all, we found out that beside the necessity to facilitate the SEAr-
like step, there are some other factors affecting the eventual
product formation. It remains unclear why 12 does not undergo
the second HF-elimination while 14 readily reacts, since the
latter is supposed to be less prone to undergo the electrophilic
attacks. We have tested several precursors that might lead to
1 via three consequent C-C bond formations and found that 8
indeed gives 1 on activated alumina, which makes the reaction
attractive since only few steps exploiting conventional methods
are required in order to obtain 8. Moreover, the experimental
data provide us with some valuable insights into the design of
the precursors for other related buckybowls.
Our further endeavors were aimed to manage the formation of
the maximum number of C-C bonds at the last stage.
According to this strategy, we have attempted to implement the
synthetic scheme (Figure 4a) where 1 was supposed to be
formed from 16 after the sequence of four aryl-aryl couplings.
Unfortunately, our efforts led neither to the desired product nor
to any of the expected intermediate products, nevertheless the
initial reagent was not observed in the extract after the
reaction. The observed experimental data evidence that the
attack of the formed cation-like species does not occur and
some other side reactions proceed leading to the irreversible
adsorption of the by-products on the solid phase. We assume
that the reaction can be hindered by two possible factors. The
unfavorable rigid geometry and thus the lack of the
conformational flexibility, which plays the crucial role in the
intramolecular interactions, serve as the first factor. Secondly,
the reactive double C-C bond of 16 might be polarized on
frustrated Lewis pairs formed on the surface of activated
alumina, thus leading to undesired side reactions.
In order to avoid the undesired reactions of the double bond
and to increase the number of possible conformers, we
modified our approach as shown in Figure 4b. The
incorporation of sp3-hybridized carbon atoms in 17 introduces
additional flexibility into a system circumventing substantial
strain energy arising during aryl-aryl coupling otherwise. The
required precursor 18 can be obtained by standard McMurry
coupling of the respective fluorinated benzophenone.[28] Thus
we have suggested a possible pathway enabling facile three-
step synthesis of 1 from 19. After subjecting 19 to the
conditions of the alumina mediated HF-elimination we
observed rather complicated mixture, nevertheless we
managed to isolate three major fractions exploiting HPLC.
Surprisingly, instead of the expected intermediate 17 we have
observed the formation of the target molecule 1 along with
several interesting side products such as 20, 21 and 22. At this
point, multiple fundamental questions arise, first of all the major
question is how the formation of the middle double bond
occurs. Therefore, we have investigated whether
dehydrogenation may occur under the conditions of HF-
elimination. For this purpose, we subjected 9H,9'H-9,9'-
bifluorene to activated alumina under vacuum and observed a
significant conversion to 9,9'-bifluorenylidene which was
detected via HPLC analysis. This transformation shows that
alumina can initiate the dehydrogenation. Moreover, in
accordance with the previous experiment (Figure 4a) the
dehydrogenation step must occur no sooner as at least one of
C(aryl)-C(aryl) bond is formed. Otherwise, it leads to 16 and
the abruption of the further reaction. One of the possible
explanations for the presence of 20 is supported by the
formation of the cation-like species, which can hardly undergo
intramolecular SEAr due to the presence of the electron-
withdrawing groups. The system may undergo several
alternative reactions e.g. such as a double bond formation via
loss of HF (see SI), thus three out of four fluorine atoms are
responsible for aryl-aryl coupling while the elimination of the
fourth HF leads to the double bond formation. The described
above processes may serve as an account for the formation of
both 1 and 20, however the mechanism leading to the
formation of 21 and 22 remains obscure. The ipso-substitution
appears to be the only reasonable explanation of the
observations, which accounts for the positions of the fluorine
atoms in 21 and 22 (see SI). Since we consider the SEAr to be
the direction-determining step, all three processes described
above as side reactions might be suppressed upon the
facilitation of SEAr. It comes naturally to promote the
electrophilic attack via disposal of fluorine atoms from the π-
system. For this reason, we have obtained 23, which is
supposed to undergo SEAr easily, and could also shed a light
on the sequence of couplings. The following experiment
(Figure 4d) revealed two main products 24 and 25, which
correspond to double HF and H2 elimination. In addition, by
means of HPLC we have found some phenanthrene
derivatives that were not isolated or characterized though,
however the collected data allows us to conclude that at first
the
six-membered
ring
is
formed
leading
to
Figure 4. Multi-assemblies of fluorinated tetraphenyl ethanes.
diphenylphenanthrene derivatives, after which multiple
pathways are possible. Having these observations in hand we
tried to carry out the synthesis of 1 in the most efficient way.
This article is protected by copyright. All rights reserved.