α-Dicationic Chelating Phosphines: Synthesis and Application to the

Mar 21, 2017 - α‑Dicationic Chelating Phosphines: Synthesis and Application to the ..... (b) Robertson, A. P. M.; Gray, P. A.; Burford, N. Angew. C...
1 downloads 3 Views 3MB Size
Article pubs.acs.org/JACS

α‑Dicationic Chelating Phosphines: Synthesis and Application to the Hydroarylation of Dienes Lianghu Gu,† Lawrence M. Wolf,‡ Adam Zieliński,† Walter Thiel,§ and Manuel Alcarazo*,† †

Institut für Organische und Biomolekulare Chemie, Georg-August-Universität Göttingen, Tammannstraße 2, 37077 Göttingen, Germany ‡ Department of Chemistry, University of Massachusetts Lowel, Lowell, Massachusetts 01854, United States § Max-Planck-Institut für Kohlenforschung, Kaiser-Wilhelm-Platz 1, 45470 Mülheim an der Ruhr, Germany S Supporting Information *

ABSTRACT: A series of new P^P-chelating ligands constituted by a dicationic −[P(H2Im)2]+2 unit (H2Im = 1,3-dimethyl-4,5-dihydroimidazol-2ylidene) and a −PPh2 group connected through structurally different backbones have been synthesized. Evaluation of their reactivity toward different metal centers provides evidence that the dicationic fragment, otherwise reluctant to coordinate metals, readily participates in the formation of chelates when embedded into such a scaffold. Moreover, it significantly enhances the Lewis acidity of the metals to which it coordinates. This property has been used to develop a Rh catalyst that efficiently triggers the hydroarylation of dienes with electron-rich aromatic molecules. Kinetic studies and deuterium-labeling experiments, as well as density functional theory calculations, were performed in order to rationalize these findings.



R2P^P(L2)+2, evaluate the influence of the positive charges on their donor properties by experimental and theoretical methods, and provide a preliminary demonstration of their beneficial effects in catalysis.

INTRODUCTION The strategic placement of one or more positively charged group(s) directly attached to the central phosphorus atom of phosphines provides a useful entry, complementary to polyfluorination, for the design of ligands of enhanced πacceptor character. In addition, the absence of labile P−halogen bonds in these α-cationic phosphines significantly enhances their robustness and facilitates their handling.1 Making use of these distinctive attributes, we recently developed a series of new Au(I) and Pt(II) catalysts where the natural π-acidity of these metal centers is substantially increased and, as a result, an unmatched activation of alkynes toward nucleophilic attack is achieved.2 Unfortunately, the use of cationic ligands also faces an important shortcoming: the strongest π-acceptor phosphines that might be conceived, namely, di- and tricationic ones, show very little propensity to form coordination complexes. Only Pt and Au derivatives are known.3 In an attempt to overcome this intrinsic shortcoming and to expand the use of polycationic ligands to other transformations, which might also benefit from the unique electronic properties that they impart, we envisioned the introduction of an additional donating atom in the ligand architecture as an anchor that should bring the cationic moiety into the vicinity of the metal and thus induce its coordination. This strategy was applied with success in the past to coordinate groups that were reluctant to do so, such as monocationic nitrenium moieties or boranes and their heavier analogues.4,5 In the following, we describe the synthesis and coordination chemistry of several dicationic chelating phosphines of the general structure © XXXX American Chemical Society



RESULTS AND DISCUSSION

Synthesis of Dicationic Chelating Phosphines. At the beginning of our investigation, we decided to prepare diphosphines 1 and 2, both containing a neutral −PPh2 moiety and a strong acceptor dicationic [−P(H2Im)2]+2 group but separated by structurally different linkers. In 1, the adjacent position of both phosphorus atoms ideally favors their simultaneous coordination to the incoming metal, whereas in 2, the more flexible 2,2′-biphenylene spacer relaxes geometric restrictions and serves as a suitable model for future chiral versions of these systems. Compounds 1a,b were obtained in good yields through a two-step sequence consisting of the condensation of known diphosphine 3 with commercially available 2-chloro-1,3-(dimethyl)imidazolinium chloride 4, followed by anion exchange (Scheme 1).6 This same synthetic route was satisfactorily extended to the preparation of 2 once the necessary primary phosphine 5 was made available from the already described iodide 6.7 The 31P NMR spectrum of both compounds consists of two doublets (δP = −8.5 (P2), −43.8 (P1) ppm, JPP = 212 Hz for 1; and δP = −14.7 (P2), −44.1 (P1) ppm, JPP = 86.7 Hz for 2), Received: February 10, 2017

A

DOI: 10.1021/jacs.7b01441 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article

Journal of the American Chemical Society

noteworthy that the HOMO−LUMO gap for 1 is 1.0 eV lower than that of the neutral analogue 7. In addition, AIM9 and NBO10 analysis support the presence of a weak P−P bond: the electron density and its corresponding Laplacian reveal a closed-shell dative interaction with a Wiberg bond index of 0.124. Remarkably, this interaction cannot be attributed purely to the proximity induced by the bridging ligand backbone; when the phenyl backbone was replaced by two hydrogen atoms, a constraint-free optimization of the fragments resulted in a structure (1-frag) in which the computed P−P distance was only slightly larger (3.13 Å) than that in 1 (2.99 Å, Table 1). The M06-L/TZVP binding energy in this intermolecular

Scheme 1. Synthesis and Molecular Structures of 1 and 2 in the Solid Statea

Table 1. Geometries and Bonding Analysis for 1 and 1-fraga

ρb(P1−P2) ∇2ρb(P1−P2) BO nP1→σ*P1−C1

1

1-frag

0.033 0.036 0.124 6.7

0.024 0.030 0.155 8.6

1-frag ΔEPauli ΔEelstat ΔEorb ΔEdisp ΔEint ΔEdist ΔE

34.2 −26.4 −29.6 −18.0 −39.8 3.5 −36.3

Left: AIM parameters ρb and ∇2ρb, Wiberg bond index BO, and NBO perturbation energies (kcal/mol) for 1 and 1-frag. Right: Energy decomposition analysis (kcal/mol) for 1-frag. Results from BP86-D3/ TZ2P//M06-L/TZVP calculations.

a

a

Hydrogen atoms, anions, and solvent molecules were removed for clarity; ellipsoids are set at 50% probability.8 Reagents and conditions: (a) Et3N (2 equiv), THF, 60 °C; then NaSbF6 (excess) in CH3CN, 72%; (b) n-BuLi, ClP(O)(OEt)2, 47%; (c) LiAlH4, 85%; (d) 1 (2 equiv), Et3N (2 equiv); then NaSbF6 or NaBF4 (excess) (1a, 72%; 1b, 65%).

Lewis adduct based on P-donor/P-acceptor fragments was found to be −31.3 kcal/mol.11 Finally, evaluation of the physical nature of this interaction through energy decomposition analysis12 at the BP86/TZ2P//M06-L/TZVP level indicated that the most prominent contributions to P−P bonding come from orbital interactions (ΔEorb; 40%), which are stronger than the electrostatic and dispersion interactions (see Table 1 and Supporting Information). The presence of this capto-dative interaction already in the free ligand suggests that the [P(H2Im)2]+2 moiety should behave as an excellent acceptor provided that its coordination to metals can be induced. Evaluation of Electronic Properties. The global donor ability of 1 was estimated through comparative analysis of the CO stretching frequencies in the Mo complexes 8−11 (Figure 2). The νCO values recorded for dicationic 8 (2043, 1973, 1938 cm−1) are significantly higher than those observed for the other Mo(0) pincer complexes tested, thus confirming the exceptionally high acceptor strength of the [P(H2Im)2]+2 unit, which surpasses that of strong π-acceptor groups such as −P(C6F5)2 or even −P(1-pyrrole)2 by a wide margin. Moreover, the isolation of 8 constitutes the first experimental evidence of the capacity of 1 to act as a chelate in coordination complexes. Cyclic voltammetry studies also support this conclusion. Ligands 1 and 2 gave oxidation peaks at 1.050 and 0.974 V, respectively, which can be assigned to the oxidation of the −PPh2 groups. Comparison with reference compounds reveals that the −PPh2 lone pairs in 1 and 2 are more internal than those in PPh3 (Table 2, entry 3), probably as a result of the

which can be attributed with certainty to the −PPh2 and [−P(H2Im)2]+2 moieties, respectively. Single crystals suitable for X-ray structure determination of both compounds were obtained, and their ORTEP diagrams are depicted in Scheme 1. Of most interest is the P1−P2 distance (3.076 Å) in 1, which is significantly shorter than that in its neutral analogue 7 (3.165 Å), suggesting the possibility of a Ph2P→ [P(H2Im)2]+2 dative interaction in the free ligand as result of the strong Lewis acid character generated on P2 by the two positively charged substituents. This hypothesis was confirmed computationally. Inspection of the frontier orbitals at the M06-L/TZVP level indicated that the HOMO is largely localized on the neutral phosphorus atom, while the LUMO and LUMO+1 orbitals are located mainly on the dicationic fragment (see Figure 1). It is also

Figure 1. Frontier molecular orbitals of 1 and their energies (in eV) computed at the M06-L/TZVP level. B

DOI: 10.1021/jacs.7b01441 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article

Journal of the American Chemical Society

both [−P(H2Im)2]+2 groups adopt a distal orientation (see the Supporting Information for its X-ray structure). Interestingly, although 14 was isolated as a dimer in the solid state, the Coulomb repulsion between the components will surely weaken the strength of the chloride bridges in solution. This is evident from the trapping of monomeric 15 after addition of 2 equiv of acetonitrile to a dichloromethane solution of 14 (Scheme 2). Analogously, reaction of 2 with [RhCl(cod)]2 in Scheme 2. Syntheses and X-ray Structures of 14−17a

Figure 2. Evaluation of the donor ability and X-ray structure of 8. Hydrogen atoms, anions, and solvent molecules were removed for clarity; ellipsoids are set at 50% probability.8 Wavenumbers in cm−1.

Table 2. Electrochemical Redox Potentials

a

Hydrogen atoms, anions, and solvent molecules were removed for clarity; ellipsoids are set at 50% probability.8 Reagents and conditions: (a) [RhCl(cod)]2 (0.5 equiv), 92%; (b) CH3CN (2 equiv); (c) [RhCl(cod)]2 (0.5 equiv) in CH3CN, 93%; (d) PtI2(PhCN)2, 95%. a

Oxidation/reduction potential peaks reported in V and calibrated versus ferrocene/ferrocenium, Bu4NPF6 (0.1 M) in CH2Cl2. bNo reduction signal observed from 0 to −2.5 V. cNo signal detected from 0 to 2 V.

acetonitrile afforded the desired Rh complex 16, in which a biphenyl group serves as the backbone that ties together the two phosphines. Ligand 1 also demonstrated its ability to form Pt derivatives. Addition of PtI2(PhCN)2 to DCM solutions of 1 afforded a yellow solid 17 whose 31P NMR spectrum showed signals (δP = 39.2 (P2), 5.4 (P1) ppm; JPP = 4.6 Hz) shifted downfield compared to the original resonances and the appearance of characteristic 195Pt satellites (1JP(2)Pt = 2854 Hz; 1JP(1)Pt = 3457 Hz) which indicate coordination of both P centers to Pt. The connectivities of 16 and 17 could be unambiguously confirmed by X-ray analysis. Catalysis. With complexes 14−16 in hand, we initiated a catalysis study by revisiting the previously reported Rh(I)catalyzed hydroarylation of phenyl-1,3-butadiene 18a with indole (Chart 1).13 The kinetics of this benchmark transformation have been shown to be remarkably sensitive to the relative Lewis acidity at the Rh center in the catalyst; thus, it seems to be an appropriate model to test the impact of our new ligands.13c During our first assay, we observed the formation of 19a in moderate yields in the presence of 5 mol % of 15 in dichloroethane at 70 °C. Further optimization of the reaction conditions revealed that the use of 10 mol % of KB(ArF)4 (potassium tetrakis(pentafluorophenyl)borate) is crucial. This

strong electron-withdrawing effect of the dicationic group, which is markedly stronger in 1 due to the closer proximity of the two phosphorus atoms imposed by the o-phenylene linker. The reduction potentials of 1 (−1.528 V) and 2 (−1.504 V) that reflect the electrophilic character of the [−P(H2Im)2]+2 fragment are slightly larger (in absolute magnitude) than that of 13 (Table 2, entry 6), implying a marginal loss in acceptor strength compared with the monodentate analogue. The bonding Ph2P→[P(H2Im)2]+2 interaction (described above) may effectively account for the changes in the observed reduction and oxidation potentials when going from 1 to PPh3 and 13 (Table 2, entries 3 and 6), by considering the mixing of the lone-pair orbital on the neutral phosphorus with antibonding orbitals on the dicationic moiety. Coordination Studies. Encouraged by these results, we decided to evaluate the ability of 1 and 2 to coordinate different metals centers. Specifically, given our long-standing interest in hydroarylation reactions, we analyzed their reactivity toward Rh(I) and Pt(II) centers. Salt 1 readily displaces cyclooctadiene in [RhCl(cod)]2, generating a dinuclear tetracation 14 in which C

DOI: 10.1021/jacs.7b01441 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article

Journal of the American Chemical Society Chart 1. Preliminary Scope of the Hydroarylation of Dienesa

engaged in the reaction, though in the latter case the yield was significantly lower, probably due to the relatively low nucleophilicity that is characteristic of this system. Identical results were observed when these experiments were carried out using a 2:1 mixture of 1 and [RhCl(CO)2]2 instead of complex 15.15 It is worth mentioning that less than 5% conversion of 18a to 19a was observed when 1,2-bis(diphenylphosphanyl)benzene 7 or the stronger acceptors 1-(diphenyl-phosphanyl)-2-bis(pentafluorophenyl)phosphanylbenzene 12, 1-(diphenylphosphanyl)-2-di(1-pyrroyl)phosphanylbenzene 20, and even 2 were used as ancillary ligands under otherwise identical reaction conditions. The unsatisfactory results obtained with 2 are probably due to the hemilability of this ligand under the reaction conditions, which is a consequence of the more flexible bisphenylene backbone. Density Functional Theory Calculations on the Reaction Mechanism. To better understand the origin of the enhanced reactivity observed, we studied the mechanistic pathway of the transformation of 18a into 19a using density functional theory (DFT) calculations at the BP86-D3(CPCM)/def2-TZVP//BP86-D3/def2-SVP level (see Supporting Information for computational methodology and detailed numerical results). Two reaction pathways were explored to highlight the importance of the biscationic ligand for the reactivity observed (Figure 3). The black one includes biscationic ligand 1, whereas the red path uses a model monocationic ligand.16 Starting with the dicationic path, the reaction begins with the coordination of the substrate to the rhodium complex to form INT1 or INT2, in which the butadiene side-chain moiety adopts either a η 4 or η 2 coordination mode, respectively. Both complexes are in equilibrium, but it is the lower-energy INT2 in which only the terminal π-system is coordinated to Rh that is attacked by indole at the internal carbon of the terminal olefin to form INT3. The free energy of activation for this step is calculated to be 14.9 kcal/mol. Subsequent rotation of the molecule around the Rh−C bond forms INT4, which features a weak hydrogen bonding interaction between the chloride ligand and the hydrogen at the 3-position of the indole. Proton transfer to the Rh center takes place through TS3 to form a Rh(III) species, INT5, which undergoes a facile reductive elimination to afford the desired product 19a still coordinated through the internal CC double bond of the original butadiene moiety to Rh in INT6. Our calculations indicate that TS3 leading to the formation of INT5 is the highest-energy transition state of the complete catalytic cycle and therefore governs the kinetics of the complete process.17 Finally, product release from INT6 with association of a new substrate is only slightly uphill by 3.4 kcal/ mol. The analogous monocationic path (red) requires significantly more energy; the difference in the activation barriers between the two pathways is predicted to be 14.4 kcal/mol. Hence, it is clear that the second positive charge on the ligand is critical for sufficiently activating the diene toward nucleophilic attack of the indole. Two of the steps from the catalytic cycle deserve further discussion because they are fundamental to understanding the role of the ancillary ligand 1. The nucleophilic attack of indole to the coordinated olefin via TS1 is not an easy step, and it can only take place because of the high acceptor strength of 1. This becomes evident by comparing the reactant complex INT2 to

a

Isomeric ratios, shown in parentheses, were determined by GC-MS or 1H NMR analysis of crude reaction mixtures. bObtained as a 9:1 mixture of the 2- and 4-regiomers. cWith 10 mol % of catalyst used. All yields are isolated.

additive substantially increases the solubility of the catalyst in the reaction medium by anion exchange and thus significantly improves the obtained conversions. Under these conditions, a range of indoles with different substitution patterns were found to undergo the desired hydroarylation in good to excellent yields and regioselectivities (19a−h). These results encouraged us to test the reaction with more challenging arenes such as dimethylresorcinol derivatives, which are characterized by a significantly lower nucleophilicity parameter than indole on Mayr’s scale (N = 5.75 and 2.48 for 1-methylindole and dimethoxyresorcinol, respectively).14 Again in these cases, high conversions and excellent regioselectivities were uniformly observed (19i−k). In addition, the transformation is compatible with halogens, as demonstrated by the isolation of 19c, 19g, and 19j. Finally, azulene and benzofurane could also be D

DOI: 10.1021/jacs.7b01441 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article

Journal of the American Chemical Society

Figure 3. Gibbs free energy profile for the Rh-catalyzed hydroarylation of phenyl-1,3-butadiene with indole and precatalyst 15, calculated at the BP86-D3(CPCMDCE)/def2-TZVP//BP86-D3/def2-SVP level of DFT. Biscationic path is shown in black, and the monocationic one in red.

transfer of 0.21 and 0.08e, respectively, from the diene to the Rh complex. The presence of a significantly lower acceptor orbital in INT2 and the accumulation of positive charge in the diene fragment account for the enhanced reactivity observed for ligand 1 relative to dppbz. Also relevant is the nature of the highest-energy step, which is predicted to be the proton transfer from the substrate to Rh (via TS3), yielding an effective activation barrier of 24.6 kcal/ mol for the whole cycle. This step involves the oxidation of the metal center from Rh(I) to Rh(III) and should thus be facilitated by a strong donor rather than an acceptor ligand; however, it occurs smoothly. Hence, the acceptor strength of 1 is large enough to allow the attack of the nucleophile to the coordinated diene, a step that seems to be energetically prohibitive for the other ligands tested but not too large to hinder the subsequent metal protonation. Finally, the experimentally observed primary kinetic isotope effect of 2.1 is consistent with a proton transfer during the ratedetermining step occurring through a nonlinear transition state; thus, it additionally supports our mechanistic proposal in which TS3 is rate-determining.

the analogous intermediates in which 1 has been formally exchanged with neutral 7 (INT2n) and the model monocationic ligand INT2′ (Figure 4). In the addition step, the

Figure 4. Relevant unoccupied orbitals of INT2 (right), INT2′ (middle), and the neutral analogue INT2n (left). The orbital energies (ε) and the NBO charge transfers (Δ) are also given.



CONCLUSIONS We present the first synthesis of bidentate α-dicationic phosphines containing one dicationic and one neutral phosphine unit connected by different backbones. This architecture induces the coordination to metals of the dicationic group, which otherwise is reluctant to react. Infrared studies as well as theoretical calculations indicate that once these ligands are in the coordination sphere of the desired metal, they show pronounced acceptor characteristics which surpass those of their neutral counterparts. Making use of these distinct properties, we have developed a Rh catalyst that shows

leading orbital interaction with the nucleophile will involve the π*-system of the diene, which is the LUMO in INT2n and INT2′ and the LUMO+3 in INT2 (where the LUMO, LUMO +1, and LUMO+2 orbitals are localized elsewhere but are all similar in energy). The LUMO+3 orbital of INT2 is significantly lower in energy than the LUMOs of both neutral INT2n (by ∼5 eV) monocationic INT2′ (by ∼2 eV). Furthermore, the computed NBO charge transfer upon complexation reveals a net donation of 0.02e from the metal to the diene in the neutral complex, whereas complexation in INT2 and INT2′ results in a net E

DOI: 10.1021/jacs.7b01441 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Article

Journal of the American Chemical Society

(3) For the few known examples, see ref 1d and (a) Coles, M. P.; Hitchcock, P. B. Chem. Commun. 2007, 5229−5231. (b) Petuškova, J.; Patil, M.; Holle, S.; Lehmann, C. W.; Thiel, W.; Alcarazo, M. J. Am. Chem. Soc. 2011, 133, 20758−20760. (4) (a) Tulchinsky, Y.; Iron, M. A.; Botoshansky, M.; Gandelman, M. Nat. Chem. 2011, 3, 525−531. (b) Tulchinsky, Y.; Kozuch, S.; Saha, P.; Botoshansky, M.; Shimon, L. J. W.; Gandelman, M. Chem. Sci. 2014, 5, 1305−1311. (c) Bouhadir, G.; Amgoune, A.; Bourissou, D. Adv. Organomet. Chem. 2010, 58, 1−106. (d) Ke, I.-S.; Gabbaï, F. P. Inorg. Chem. 2013, 52, 7145−7151. (e) Ke, I.-S.; Jones, J. S.; Gabbaï, F. P. Angew. Chem., Int. Ed. 2014, 53, 2633−2637. (5) Monocationic chelating ligands have been previously described; however, the cationic phosphines employed are also known to act as monodentate ligands on their own. See: (a) Abdellah, I.; Lepetit, C.; Canac, Y.; Duhayon, C.; Chauvin, R. Chem. - Eur. J. 2010, 16, 13095− 13108. (b) Canac, Y.; Debono, N.; Vendier, L.; Chauvin, R. Inorg. Chem. 2009, 48, 5562−5568. (c) Ruíz, J.; Mesa, A. F. Chem. - Eur. J. 2012, 18, 4485−4488. (6) For the synthesis of 3, see: (a) Schull, T. L.; Knight, D. A. Tetrahedron: Asymmetry 1999, 10, 207−211. (b) Reetz, M. T.; Gosberg, A. Tetrahedron: Asymmetry 1999, 10, 2129−2137. (7) Cereghetti, M.; Arnold, W.; Broger, E.; Rageot, A. Tetrahedron Lett. 1996, 37, 5347−5358. (8) CCDC 1033312 (1a), 1033314 (2), 1515117 (8), 1033313 (14), 1033318 (15), 1033319 (16), 1033315 (17), and 1515296 (19n) contain the supplementary crystallographic data for this paper. These data can be obtained free of charge from The Cambridge Crystallographic data centre via www.ccdc.cam.ac.uk/data_request/cif. (9) Bader, R. F. W. Chem. Rev. 1991, 91, 893−928. (10) (a) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem. Rev. 1988, 88, 899−926. (b) Glendening, E. D.; Reed, A. E.; Carpenter, J. E.; Weinhold, F. NBO, version 3.1; University of Wisconsin: Madison, WI. (11) (a) The Lewis acid behavior of P-based species has been recently reviewed: Bayne, J. M.; Stephan, D. W. Chem. Soc. Rev. 2016, 45, 765−774. (b) Robertson, A. P. M.; Gray, P. A.; Burford, N. Angew. Chem. Int. Ed. 2014, 53, 6050−6069. (12) (a) Morokuma, K. J. J. Chem. Phys. 1971, 55, 1236−1244. (b) Ziegler, T.; Rauk, A. Inorg. Chem. 1979, 18, 1755−1759. (c) Ziegler, T.; Rauk, A. Inorg. Chem. 1979, 18, 1558−1565. (13) (a) Wang, M. Z.; Wong, M. K.; Che, C. M. Chem. - Eur. J. 2008, 14, 8353−8364. (b) Peng, J.; Du, D. M. Eur. J. Inorg. Chem. 2012, 2012, 4042−4051. (c) Roberts, C. C.; Matías, D. M.; Goldfogel, M. J.; Meek, S. J. J. Am. Chem. Soc. 2015, 137, 6488−6491. (14) Mayr, H.; Kempf, B.; Ofial, A. R. Acc. Chem. Res. 2003, 36, 66− 77. (15) Pt compound 17 did not perform satisfactorily, probably fast decomposition takes place after abstraction of one of the iodide anions with Ag[B(ArF)4]. (16) No transition state could be localized for the attack of indole to the diene substrate coordinated to [7·RhCl]. (17) Kozuch, S.; Shaik, S. Acc. Chem. Res. 2011, 44, 101−110.

remarkable activity in the hydroarylation of dienes with electron-rich hetero- and homoarenes. Ongoing research in our group is focused on the application of this ligand platform to other catalytic processes and on the development of asymmetric versions of this reaction.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.7b01441. X-ray data for 1 (CIF) X-ray data for 2 (CIF) X-ray data for 8 (CIF) X-ray data for 14 (CIF) X-ray data for 15 (CIF) X-ray data for 16 (CIF) X-ray data for 17 (CIF) X-ray data for 19n (CIF) Experimental procedures, spectral and analytical data, computational methodology, and detailed numerical results from the DFT calculations (PDF)



AUTHOR INFORMATION

Corresponding Author

*[email protected] ORCID

Walter Thiel: 0000-0001-6780-0350 Manuel Alcarazo: 0000-0002-5491-5682 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS Generous financial support from the Fonds der Chemischen Industrie (Dozentenstipendium to M.A.), the European Research Council (ERC Starting Grant to M.A.), the Deutsche Forschungsgemeinschaft (AL 1348/5-1), the Chinese Scholarship Council (doctoral fellowship to L.G.), and the DAAD (doctoral fellowship to A.Z.) is gratefully acknowledged. We are also grateful to C. Wille and H. Tinnermann for cyclic voltammetry measurements, and to Prof. C.W. Lehmann and Dr. R. Goddard (MPI-Kohlenforschung, Mülheim an der Ruhr) for solving the X-ray structures.



REFERENCES

(1) (a) Alcarazo, M. Acc. Chem. Res. 2016, 49, 1797−1805. (b) Alcarazo, M. Chem. - Eur. J. 2014, 20, 7868−7877. (c) Gaillard, S.; Renaud, J. L. Dalton Trans. 2013, 42, 7255−7270. (d) Canac, Y.; Maaliki, C.; Abdellah, I.; Chauvin, R. New J. Chem. 2012, 36, 17−27. For pioneering work on polycationic phosphines, see: (e) Weiss, R.; Engel, S. Synthesis 1991, 1077−1079. (f) Reed, R.; Réau, R.; Dahan, F.; Bertrand, G. Angew. Chem., Int. Ed. Engl. 1993, 32, 399−401. (2) (a) González-Fernández, E.; Nicholls, L. D. M.; Schaaf, L. D.; Farès, C.; Lehmann, C. W.; Alcarazo, M. J. Am. Chem. Soc. 2017, 139, 1428−1431. (b) Tinnermann, H.; Wille, C.; Alcarazo, M. Angew. Chem., Int. Ed. 2014, 53, 8732−8736. (c) Kozma, Á .; Deden, T.; Carreras, J.; Wille, C.; Petuškova, J.; Rust, J.; Alcarazo, M. Chem. - Eur. J. 2014, 20, 2208−2214. (d) Carreras, J.; Gopakumar, G.; Gu, L.; Gimeno, A.; Linowski, P.; Petuškova, J.; Thiel, W.; Alcarazo, M. J. Am. Chem. Soc. 2013, 135, 18815−18823. (e) Carreras, J.; Patil, M.; Thiel, W.; Alcarazo, M. J. Am. Chem. Soc. 2012, 134, 16753−16758. (f) Petuškova, J.; Bruns, H.; Alcarazo, M. Angew. Chem., Int. Ed. 2011, 50, 3799−3802. F

DOI: 10.1021/jacs.7b01441 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX