10.1002/anie.202004242
Angewandte Chemie International Edition
COMMUNICATION
Fe-Twe 4 also led to mixtures but resulted in an inversed
selectivity. Interestingly, the deactivated positions K3-K5 were
those preferred by the supramolecular catalyst 4, overriding the
intrinsic reactivity of the substrate (Table 1, entry 1 vs 2).
substrate to the tweezer (see Supporting Information, p. S101-
107). According to the calculations, the two binding modes are
relatively close in energy indicating that C6-9 oxidation not only
stems from a background reaction but also from binding mode 1.
However, binding mode 2 (the cavity is filled with a acetonitrile
solvent, not shown in Fig. 3) is preferred by approx. 5 kJ/mol, in
accordance with the experimental results. Since binding mode 2
depends on the guest molecule acetonitrile, the observed
selectivity for K3-5 should be solvent dependent. Indeed, the
selectivity is greatly reduced with trifluoroethanol, and disappears
completely with the larger hexafluoro isopropanol as solvent (see
Supporting information, p. S35-36). These results provide further
evidence that the observed oxidation of the unactivated positions
C3-5 stem from substrate binding to the tweezer moiety of catalyst
4.
Several control experiments were carried out to elucidate
the role of the supramolecular recognition motif. First an
experiment in which the two parts of Fe-Twe 4 were added as
separate entities (tweezer 8b (5 mol%), and Fe-Br 2 (5 mol%))
was performed (Table 1, entry 4). The selectivity was significantly
reduced and similar to the results of Fe-Br 2, demonstrating that
the tweezer has to be covalently linked to the oxidation catalyst to
achieve high selectivity. In separate experiments, we tried to
reduce the binding ability of the substrate via methylation of the
+
amine residue. The oxidation of C10-NMeH2 already delivered
reduced selectivities (entry 5 vs 6) while with dimethylated C10-
NMe2H+ as substrate, the selectivity was almost completely lost.
These results strongly indicate that the substrate binds to the
tweezer via hydrogen bonds. The yields in these two cases
(entries 6 and 8) were only slightly reduced in comparison to entry
2 which suggest that oxidation without specific binding to the
tweezer is taking place as a background reaction. This is also
indicated by the oxidation of cyclohexane by both catalysts (see
Supporting Information, p. S37). In a competition experiment,
decylammonium and cyclohexane were subjected to the oxidation
reactions with Fe-Br 2 and Fe-Twe 4 in equal amounts which
resulted in only slightly increased selectivity for decylammonium
with Fe-Twe 4. The background reaction was much less
pronounced with 3,[11] presumably due to the oxidant being
blocked from two sides by the crown ether moieties. A third series
of control experiments was performed with the aim of inhibiting
substrate binding inside Fe-Twe 4. As Inhibitors, NH4PF4, NaOTf
and methyl viologen dichloride hydrate were explored. The yields
of the oxidation products, as well as the selectivity for C3-4
decreased. However, these results are difficult to interpret since
the inhibitor, also inhibits oxidation of the regular catalyst Fe-Br 2
devoid of a tweezer moiety. However, due to the reduced
selectivity these experiments also indicate some background
oxidation with regular “solution” selectivity at C6-9.
In summary, we reported the synthesis of a supramolecular
oxidation catalyst capable of overriding the intrinsic reactivity in
aliphatic C–H oxidation of alkyl ammonium salts. The main
products formed were ketones at carbons C3 and C4, positions
that are intrinsically strongly deactivated and therefore not formed
to a significant degree with other catalysts. Although the
selectivities have clearly to be improved to achieve synthetically
useful yields, these results augur well for the selective oxidation
of unactivated C–H positions on complex carbon frameworks.
Acknowledgements
This work was supported by the Swiss National Science
Foundation as part of the NCCR Molecular Systems Engineering
program. The authors thank Fabian Bissegger for VT-NMR
experiments, Dr. Michael Pfeffer for HR-MS analysis, and Dr.
Joan Serrano Plana for helpful discussions. Calculations were
carried out on the ETH Euler cluster.
Keywords: C–H oxidation • molecular recognition •
Subsequently, we studied the oxidation of several aliphatic
ammonium salts with different chain lengths (Table 1, Figure 3).
For all oxidation reactions with Fe-Twe 4 a pronounced selectivity
increase for the C3–C4 positions was observed compared to the
non-directed oxidations. In fact, with most substrates, ketones K3
or K4 were the favored products for Fe-Twe 4 oxidation reactions.
The yields, however, were generally lower for catalyst 4, which is
presumably due to catalyst decomposition during the oxidation
reaction (see Supporting Information, p. S39-42). The only
exception is the oxidation of C7-NH3+ in which almost all positions
are deactivated.[16] Moreover, substrates with longer alkyl chains
mostly resulted in higher yields compared to the short ones, a
trend also observed with 3.[11] Regarding the selectivities, in
principle, two different binding motifs can be envisioned for
catalyst 4 (Figure 3b): (1) The binding of the aliphatic chain inside
the cavity of the tweezer. This binding mode is observed in
aqueous solution, presumably due to the hydrophobic effect.[14] It
would expose positions C6-C8 to the oxidant. (2) Without the
hydrophobic effect, the sole binding to the polar end groups (urea
carbonyl and methoxy oxygen) of the tweezer would be feasible,
favoring the oxidation of positions C3-C5. The oxidation results
obtained clearly suggest that the second binding mode is the
predominant one. Molecular modeling was performed to
investigate the suggested binding modes of the ammonium
regioselectivity • catalysis • supramolecular chemistry
[1]
[2]
a) M. C. White, J. Zhao, J. Am. Chem. Soc. 2018, 140, 13988-14009; b)
W. R. Gutekunst, P. S. Baran, Chem. Soc. Rev. 2011, 40, 1976-1991.
a) T. Brückl, R. D. Baxter, Y. Ishihara, P. S. Baran, Acc. Chem. Res.
2012, 45, 826-839; b) J. F. Hartwig, M. A. Larsen, ACS Cent. Sci. 2016,
2, 281-292; c) T. Newhouse, P. S. Baran, Angew. Chem. Int. Ed. 2011,
50, 3362-3374; d) M. S. Chen, M. C. White, Science 2010, 327, 566-
571; e) P. E. Gormisky, M. C. White, J. Am. Chem. Soc. 2013, 135,
14052-14055; f) K. Chen, J. M. Richter, P. S. Baran, J. Am. Chem. Soc.
2008, 130, 7247-7249; g) B. Li, M. Driess, J. F. Hartwig, J. Am. Chem.
Soc. 2014, 136, 6586-6589; h) M. A. Bigi, S. A. Reed, M. C. White, J.
Am. Chem. Soc. 2012, 134, 9721-9726.
[3]
[4]
a) P. R. Ortiz de Montellano, Chem. Rev. 2010, 110, 932-948; b) R.
Fasan, ACS Catal. 2012, 2, 647-666.
a) P. Dydio, J. N. Reek, Chem. Sci. 2014, 5, 2135-2145; b) R. Breslow,
S. D. Dong, Chem. Rev. 1998, 98, 1997-2012; c) S. Das, G. W.
Brudvig, R. H. Crabtree, Chem. Commun. 2008, 413-424; d) D. Vidal,
G. Olivo, M. Costas, Chem. Eur. J. 2018, 24, 5042-5054; e) F. Burg, T.
Bach, J. Org. Chem. 2019, 84, 8815-8836; f) N. R. Mote, S. H. Chikkali,
Chem. Asian J. 2018, 13, 3623-3646; g) M. Raynal, P. Ballester, A.
Vidal-Ferran, P. W. van Leeuwen, Chem. Soc. Rev. 2014, 43, 1660-
1733; h) S. Carboni, C. Gennari, L. Pignataro, U. Piarulli, Dalton Trans.
2011, 40, 4355-4373.
[5]
examples of similar approaches concerning C-H borylation:a) Y.
Kuninobu, H. Ida, M. Nishi, M. Kanai, Nat. Chem. 2015, 7, 712; b) H. J.
Davis, M. T. Mihai, R. J. Phipps, J. Am. Chem. Soc. 2016, 138, 12759-
12762; c) M. E. Hoque, R. Bisht, C. Haldar, B. Chattopadhyay, J. Am.
Chem. Soc. 2017, 139, 7745-7748; examples of similar approaches
concerning olefin functionalization: d) P. Thordarson, E. J. Bijsterveld,
A. E. Rowan, R. J. Nolte, Nature 2003, 424, 915-918; e) T. Šmejkal, B.
Breit, Angew. Chem. Int. Ed. 2008, 47, 311-315; f) P. Dydio, W. I. Dzik,
4
This article is protected by copyright. All rights reserved.