Angewandte
Communications
Chemie
molecular design.[11] Specifically, the trifluoro motif resulting
importance of Brønsted acidity in ArIX2 mediated proces-
ses,[15d,18] variation of the amine:HF ratio was explored using
mixtures of commercially available NEt3:HF 1:3 and pyr:HF
1:9.2 (Olahꢀs reagent). Initial attempts to induce difluorina-
tive ring expansion with an amine:HF ratio of 1:3 generated
the desired geminal fluorinated tetralin 9 in only 7% yield
(Table 1, entry 1). However, adjusting the ratio to 1:4.5 led to
a significant enhancement in efficiency (61%, entry 2).
Further increasing the amine:HF ratio to 1:6 proved to be
detrimental (47%, entry 3) and thus the remainder of the
study was conducted with amine:HF 1:4.5. A solvent screen
(entries 4–9) identified DCE as being the optimal reaction
medium for the title transformation (74% yield, entry 5).
Fluorinated solvents such as hexafluoroisopropanol (HFIP)
and ethyl trifluoroacetate (ETFA) led to moderate yields
(48% and 56%, respectively, entries 6–7). Furthermore, non-
halogenated solvents proved to be detrimental (entries 8–9).
Replacing Selectfluor with m-CPBA (entry 10) or oxone
(entry 11) did not lead to an improvement, nor did increasing
or decreasing temperature (entries 12 and 13). Lowering the
catalyst loading led to a slight decrease in yield, and the
control experiment without p-TolI supports the involvement
of an I(I)/I(III) catalysis paradigm (54% and < 5%, respec-
tively, entries 14–15).
Having established suitable reaction conditions for the
difluorinative ring expansion (Table 1, entry 5) the scope and
the limitations of the transformation were explored
(Scheme 1). Initially, the parent scaffold 9 was prepared
under the standard reaction conditions, and the process could
be scaled up to 1 mmol without loss of catalytic efficiency.
Halogens were found to be compatible with this protocol as is
exemplified by products 10–14 (up to > 95% yield), enabling
the regioisomeric bromides 12–14 to be prepared as syntheti-
cally-versatile coupling partners for downstream manipula-
tion. Although it was possible to generate the methyl-
derivative 15, the disparity in yield when compared with
electron deficient systems prompted a more detailed Ham-
mett analysis (vide infra). Electron-deficient substrates
proved to be highly competent precursors as exemplified by
the triflate (16), trifluoromethyl (17) and cyano (18) species
(up to 92% yield). Phthalimide 19 was smoothly generated to
provide access to masked aniline derivatives, and a substrate
with a pendant a,b-unsaturated ester (20) demonstrates the
chemoselectivity of the transformation. The addition of
substituents on the saturated ring system was tolerated (21
and 22), and catalysis enabled the formation of the tetra-
fluorinated compound 23 and mixed halogen system 24. In all
cases, reactions performed in the absence of p-TolI led to
< 5% yield for 9–24. To correlate the electronic signature of
the aryl fragment with catalysis efficiency, the Hammett
values sp and sm of several electronically diverse compounds
were plotted against the 19F NMR yields (Scheme 1, bottom).
The plot underscores the fact that strong electron-withdraw-
ing groups on the aryl fragment facilitate difluorinative ring
expansion.
ꢀ
from isosteric replacement of the symmetric [CH2 CH2]
ꢀ
motif distal to the aryl ring by [CF2 CHF] would circumvent
the conformational lability intrinsic to the parent tetralin
scaffold: the introduction of hyperconjugative interactions
(sCꢀH!sCꢀF*) would render the two half chairs non-degener-
ate. To access this novel class of fluorinated heterocycles,
fluorinated methyleneindanes (8) were selected as substrates.
[12]
It was envisaged that exposure to in situ generated p-TolIF2
under the auspices of I(I)/I(III) catalysis[13,14,15] would induce
a fluorinative ring expansion via an ephemeral, tricyclic
phenonium ion[16] to liberate the desired product (9)
(Figure 1, bottom).
Confidence in the feasibility of this catalysis-based
strategy stemmed from a plenum of stoichiometric ring
expansion processes. Pertinent examples include the gener-
ation of difluoro ethers from aryl-substituted ketones using
XeF2 by Zupan and co-workers.[17] Furthermore, the ability of
hypervalent iodine reagents to induce ring expansion with Pd
and Cu or BF3·OEt2, has been elegantly demonstrated by the
groups of Szabꢂ[14e] and Murphy,[14g,j] respectively. To devise
a catalysis-based platform to access novel, trifluorinated
tetralins, 2-fluoro-methyleneindane (8) was selected as
a model substrate for reaction optimisation (Table 1). It was
envisaged that this allyl fluoride would engage with p-TolIF2,
generated by in situ oxidation from inexpensive p-TolI, to
forge the desired carbocycle 9, where the CHF unit would
function as a conformational control unit.
Initially, the transformation was investigated using Select-
fluor as the terminal oxidant in CHCl3 at ambient temper-
ature using HF as a convenient fluoride source. Given the
Table 1: Optimisation of the reaction conditions.
entry[a]
oxidant
amine:HF[b]
solvent
yield[c]
1
2
3
4
5
6
7
8
9
10
11
12[d]
13[e]
14[f]
15[g]
Selectfluor
Selectfluor
Selectfluor
Selectfluor
Selectfluor
Selectfluor
Selectfluor
Selectfluor
Selectfluor
m-CPBA
1:3.0
1:4.5
1:6.0
1:4.5
1:4.5
1:4.5
1:4.5
1:4.5
1:4.5
1:4.5
1:4.5
1:4.5
1:4.5
1:4.5
1:4.5
CHCl3
CHCl3
CHCl3
CH2Cl2
DCE
HFIP
ETFA
toluene
CH3CN
DCE
7%
61%
47%
71%
74%
48%
56%
54%
28%
63%
39%
65%
29%
54%
<5%
Oxone
DCE
DCE
DCE
DCE
Selectfluor
Selectfluor
Selectfluor
Selectfluor
DCE
[a] Standard reaction conditions: 8 (0.2 mmol), p-TolI (20 mol%),
oxidant (1.5 equiv.), solvent (0.5 mL), amine:HF (0.5 mL), ambient
temperature, 24 h. [b] See supporting information for the exact calcu-
lation of the amine:HF mixtures. [c] Determined by 19F NMR analysis of
the crude reaction mixture using ethyl fluoroacetate as internal standard.
[d] Reaction was performed at 508C. [e] Reaction was performed at 08C.
[f] Reaction was performed with 10 mol% catalyst. [g] Reaction was
performed without catalyst.
To confirm that electron-rich groups suppress catalysis
(see Scheme 1, lower), the difluorinative ring expansion to
generate 25 was attempted, but led to degradation of the
starting material (Scheme 2). Furthermore, the requirement
2
ꢀ 2021 The Authors. Angewandte Chemie International Edition published by Wiley-VCH GmbH
Angew. Chem. Int. Ed. 2021, 60, 1 – 6
These are not the final page numbers!