10.1002/anie.201907359
Angewandte Chemie International Edition
COMMUNICATION
Table 3. Relative volumes (VM) of CF3, SF5, TeF5, and TeF4CF3
vessels that were open to air from days to months, with virtually
no observable degradation. However, when dissolved in MeCN
and exposed to excess H2O, 1 and 13 underwent rapid and
virtually complete conversion to cis-phenyl-TeF4OH (18) and
putative eq-phenyl-TeF3(CF3)OH (19), respectively, in 5 min
(Figure 3). The 19F NMR spectrum of 18 is consistent with what
was reported by Janzen and co-workers,[31] and we can now
characterize 19 by 19F NMR: -49.99 ppm (1F, tq, J = 59.2, 28.5
Hz), -59.74 (3F, dt, J = 28.5, 19.6 Hz), -73.24 (2F, dq, J = 59.2,
19.6 Hz). In addition, we note that the 125Te NMR chemical shift
groups determined using X-ray diffraction data.[26-28]
Substituent V1/Z1-V2/Z2 (Å3)a Hirshfeld (Å3)b Promolecule (Å3)b
CF3
SF5
39.2c
62.6c
95.1c
138.1d
47.9
75.7
44.0
66.5
TeF5
92.6e
125.0
82.5e
112.2
TeF4CF3
difference between
1 (709 ppm) and its corresponding
aReferenced to unsubstituted benzophenone molecular structure to obtain initial
value for CF3, assuming a spherical hydrogen atom with VvdW = 7.2 Å3. bCalculated
using Crystal Explorer. cDetermined by comparing molecular structures with same
benzophenone core. dDetermined by comparing molecular structures with same
para-chloro-phenyl core. eDetermined for benzophenone derivative.
monohydrolysis product 18 (735 ppm) is more pronounced than
the shift difference between 13 (757 ppm) and its corresponding
monohydrolysis product 19 (762 ppm). It was also observed that
exchange of additional fluorine atoms with OH groups will
continue to happen more slowly over time.[32]
We found this quite peculiar; therefore, we conducted a short
series of control experiments. First, using styrene as a substrate,
we were able to reproduce the literature procedure (albeit
synthesizing (1,2-difluoroethyl)benzene in a significantly lower
F
F
F
F
F
F
CF3
F
Te
Te
yield
than
previously
reported).[8,9]
Notably, (1,2-
Ph
Ph
OH
OH
difluoroethyl)benzene (and all other reported products of putative
difluorination with phenyl-TeF5) would also be the expected
products of a reaction with XeF2. [29] Indeed, we confirmed that
this product is also formed in 70% yield using only stoichiometric
XeF2. We then wondered if unreacted XeF2 could initiate a chain
reaction with phenyl-TeF5, but doping a reaction mixture of 1 and
styrene with 30 mol % XeF2 only resulted in 14% yield of the
difluoride product, with no degradation of 1 by 19F NMR. In light
of the fact that product formation was only observed in reaction
mixtures where XeF2 could have been present, XeF2 may be the
real actor in difluorination under previously reported conditions.
Accordingly, the behavior of phenyl-TeF5 as a "difluorination
reagent" may be a misconception, as pure phenyl-TeF5 in our
hands was completely unreactive toward styrene and several
other olefins (as well as alkynes) under conditions outlined in the
SI. Or, at the very least, phenyl-TeF5 requires some form of
activation for such a reaction to occur; this remains unclear.
On the other hand, we were able to reproduce a number of
the reactions of phenyl-TeF5 with nucleophiles (such as
alcohols,[7] secondary amines,[7] and azides[30] – see SI for details)
that have been reported in the last few decades. Primarily cis-
isomers were formed in each case (by substitution of an
equatorial fluorine atom with a nucleophile), which is consistent
with our observation in the X-ray data that Te-Feq bonds are, on
average, longer than Te-Fax bonds. In an attempt to expand upon
these reports, we examined reactions of 1 with a variety of
additional nucleophiles (e.g. KCN, tBuNC, AgSCF3, KSCN, PhLi,
and MeLi), only to find that 1 stays completely intact. Additionally,
we explored the reactivity of 1 with a number of substrates under
300 nm irradiation, in both the presence and absence of
sensitizers. Again, in all cases, 1 is ostensibly unreactive. From
another perspective, we also examined the behavior of 1 in the
presence of TMS-X reagents (e.g. X = CF3, CF2H, CF2CF3, CN,
and acetylide) – with and without CsF – and observed no reactivity.
Lastly, we investigated the hydrolytic stability of 1 and 13. A
sample of each was stored in a borosilicate vial and kept in a
refrigerator for ca. 1 year with virtually no degradation by 19F NMR
analysis. Moreover, most of the single crystals were grown in
18
19
Figure 3. Monohydrolysis products derived from 1 (left) and 13
(right) characterized by 19F NMR.
Yet, overall, the TeF5 and TeF4CF3 groups on arenes are
surprisingly more robust than previous literature implies. Leading
up to this finding, the divergent behavior of Te-based substrates
from previously reported S- and Se-based substrates using our
newly-developed TCICA/KF oxidative fluorination approach has
enabled us 1) to synthesize several new aryl-TeF5 and aryl-
TeF4CF3 compounds, 2) to analyze their solid-state structures in
detail, and 3) to investigate (and begin to address some
misconceptions regarding) reactivity.
Given the observed
decomposition of aryl-TeF5 and aryl-TeF4CF3 compounds on
direct contact with water in solution, these compounds may less
likely be of interest to the medicinal chemistry and agrochemistry
communities; rather, their otherwise relative stability, size, and
properties may have a more realistic appeal to applications in
certain materials, such as liquid crystals.[33] Investigations are
currently ongoing in our group.
Acknowledgements
We thank the ETH transfer office for support in filing a patent
application on this work, in which C.R.P., N.S., and A.T. are listed
as inventors, as well as Dr. René Verel (ETH Zürich) for his help.
MoBiAS (ETH) is acknowledged for assistance with HRMS
analyses. Financial support was provided by ETH Zürich and the
ETH Postdoctoral Fellowship Program (C.R.P.).
Conflict of Interest
The authors declare no conflict of interest.
This article is protected by copyright. All rights reserved.