University of Groningen Redox-behavior and reactivity of formazanate ligands Mondol, Ranajit

(1)

Redox-behavior and reactivity of formazanate ligands

Mondol, Ranajit

DOI:

10.33612/diss.107969043

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.

Document Version

Publisher's PDF, also known as Version of record

Publication date: 2019

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):

Mondol, R. (2019). Redox-behavior and reactivity of formazanate ligands: Boron and aluminum chemistry. University of Groningen. https://doi.org/10.33612/diss.107969043

Copyright

Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

(2)

537515-L-sub01-bw-Mondol 537515-L-sub01-bw-Mondol 537515-L-sub01-bw-Mondol 537515-L-sub01-bw-Mondol Processed on: 21-11-2019 Processed on: 21-11-2019 Processed on: 21-11-2019

Processed on: 21-11-2019 PDF page: 37PDF page: 37PDF page: 37PDF page: 37

Chapter 2

Stable, crystalline boron complexes with

mono-, di- and trianionic formazanate

ligands

Redox-active formazanate ligands are emerging as tunable electron-reservoirs in coordination chemistry. Here we show that boron diphenyl complexes with formazanate ligands, despite their (formal) negative charge, can be further reduced by up to two electrons. A combined crystallographic, spectroscopic and computational study establishes that formazanate ligands are stable in mono-, di- and trianionic form.

This chapter has been published:

Mondol, R., Snoeken, D. A., Chang, M.-C. & Otten, E.* “Stable, crystalline boron complexes with mono-, di- and trianionic formazanate ligands” Chem. Commun., 2017, 53, 513–516.

(3)

24

2.1 Introduction

Coordination complexes containing ligand systems that are accessible in several oxidation states (‘redox-active’ ligands) are attractive for catalysis1,2_{as well as materials applications.}3

Several classes of redox-active ligands are known but those that combine a modular, straightforward synthesis with excellent chemical stability across two (or more) different redox-states remain scarce. Lappert and co-workers have reported ligand-based reduction chemistry of the popular class of β-diketiminates,4–10_{but these compounds are accessible only at very}

negative potentials (using strong reducing agents). Taking inspiration from the work by Hicks and co-workers,11_{our group has started to investigate the (redox-)chemistry of formazanate}

ligands in coordination compounds. In recent years, we12–15_{as well as others}16–22_{have taken}

advantage of the unusual chemistry and photophysics that formazanate ligands provide. In attempts to characterize in more detail the structures that result from ligand-based redox reactions in formazanate compounds, we reported previously that cyclic voltammograms of boron difluoride complexes indicate two sequential reductions take place to form the redox-series [LBF2]0/-1/-2.13 Attempts to isolate the most reduced member of this series, the dianion

[LBF2]-2, resulted in elimination of fluoride (as NaF) and formation of BN-heterocyclic

products that derive from a putative boron carbenoid intermediate.23_{In this chapter we present}

(formazanate)boron diphenyl complexes that lead to stable 2-electron reduction products which are characterized by spectroscopic and crystallographic methods.

2.2 Synthesis and characterization of boron compounds bearing

formazanate ligands in their monoanionic form

The (formazanate)boron diphenyl complexes (PhNNC(p-tolyl)NNCAr)BPh2 (Ar = Ph (2a); Ar

= Mes (2b)) are obtained upon refluxing a toluene solution containing equimolar amounts of formazan (1a or 1b) and BPh3 for 3 days. Purification by column chromatography and

recrystallization from hexane gave the products in 63% (2a) and 74% (2b) isolated yield. The

11_{B NMR spectrum shows a broad singlet at 1.74 ppm (2a) and 2.27 ppm (2b), which is similar}

to the reported (formazanate)boron difluoride analogues13_{and suggests a four-coordinated}

boron centre. Single crystals of 2a suitable for X-ray crystallography were obtained by recrystallization from hexane solution at -30 °C (Figure 2.5 and Table S2.1). Overall, the distorted tetrahedral geometry around boron is similar to that in the difluoride analogue, with the B atom displaced out of the plane defined by the 4 N atoms by 0.685 Å giving rise to crystallographically distinct B-Ph groups. However, exchange is fast on the NMR timescale and only one set of resonances is observed for the B-Ph groups in the 1_{H and}13_{C NMR spectra}

(4)

25 at room temperature. The intraligand N-N and N-C bonds are unremarkable, but the B-N bonds (1.5990(15)/1.5955(15) Å) in 2a are significantly longer than in the corresponding boron difluoride (1.5589(16)/1.5520(16) Å) due to a decrease in Lewis acidity of the boron centre. Similar to the difluoride analogue, 2a is weakly emissive in solution (THF: λmax = 505 nm, λem

= 687 nm; Stokes shift = 5246 cm-1_).

Scheme 2.1 Synthesis of (formazanate)boron diphenyl compounds (left), and cyclic voltammetry of 2a and 2b (1.5 mM solution in THF, 0.1 M [Bu4N][PF6]) recorded at 100 mVs -1_(right).

The electrochemical properties of compound 2a were established by cyclic voltammetry in THF solution (Scheme 2.1), which shows two quasi-reversible 1-electron redox-events at -1.35 and -2.26 V vs. Fc0/+_{. In comparison with the boron difluoride analogues,}13_{the reduction potentials}

for 2a are shifted to more negative potentials due to the presence of less electron-withdrawing B-substituents (Ph vs F). As anticipated, replacing an N-Ph substituent for N-Mes on the formazanate backbone (2b) results in a shift of the redox-potentials to more negative values. Moreover, the second reduction is clearly less reversible for 2b and a new oxidation wave is observed at Ep,a = -1.26 V that likely results from a chemical transformation of the initial

2-electron reduction product.

2.3 Synthesis and characterization of boron compounds bearing

formazanate ligands in their di- and trianionic forms

2.3.1 Synthesis of 1- and 2-electron reduced products

Chemical synthesis and subsequent characterization of the reduction products of 2 was attempted. The radical anions 3a/b were generated via treatment with 1 equiv of Cp*2Co and

could be isolated as green crystalline material in good yield (structure of 3b shown in Figure 2.2). More surprisingly, the reaction of compounds 2 with 2 equiv of Na/C10H8 as reducing

(5)

26

obtained in moderate yield as orange crystals of their disodium salts (4a/b) upon precipitation from THF/hexane (Scheme 2.2). While the 2-electron reduction of formazanate boron fluoride results in facile cleavage of both B-F bonds to form 2 equiv of NaF and a (transient) boron carbenoid species,23_{the B-Ph bond in 4a/b is thermally stable: when kept in THF solution at}

room temperature under an inert atmosphere no decomposition is noticeable over several days.

N N N N Ar Ph p-tol B Ph Ph 2a: Ar = Ph 2b: Ar = Mes Cp*2Co Na+C10H8 -(2 equiv) N N N N Ar Ph p-tol B Ph Ph 3a/3b Cp*2Co N N N N Ar Ph p-tol B Ph Ph 4a/4b Na Na (THF)x (THF)x THF, RT THF, RT

Scheme 2.2 Synthesis of 1-electron and 2-electron reduced (formazanate)boron diphenyl compounds.

2.3.2 Analysis of X-ray crystallographic data of 1- and 2-electron reduced

products

Single-crystal X-ray diffraction studies for compounds 3 and 4 show distorted tetrahedral geometries around the boron centre, with the ligands bound via the terminal N atoms to give 6-membered chelate rings (Figure 2.1). In comparison to the neutral precursor 2a, compounds 3 and 4 show progressive elongation of the N-N bonds (e.g., 2a: 1.3060(13)/1.3090(13) Å; 3a: 1.369(4)/1.373(4) Å; 4a: 1.428(3)/1.433(3) Å), consistent with ligand-based reduction which populates the ligand N-N π*-orbitals. At the same time, the N-C(Ar) distances shorten upon reduction, which is consistent with delocalization of electron-density into the N-Ar rings (vide infra). The crystal structure of compound 4a (Figure 2.1, middle) shows that in the solid state, the two Na+_{ions are coordinated to the dianionic boron complex via the internal N atoms of the}

formazanate ligand (closest contacts are Na-N distances of 2.351(2) and 2.390(2) Å). In contrast to 4a, the solid state structure of 4b (Figure 2.1, right) reveals a dimeric structure: the asymmetric unit contains two [(formazanate)BPh2]2- moieties that are bridged by two Na+

cations. Both of these sit in a coordination pocket formed by formazanate N-atoms and aromatic groups, so that no (for Na(1)) or only one additional THF molecule (for Na(2)) is bound to the cation. Interaction of alkali metal cations with the C(π) atoms of aromatic rings is increasingly recognized as an important structural element.24–29_{Several relatively close contacts (<2.9 Å)}

(6)

27 between Na+_{cations and aromatic carbon atoms are observed, but this does not lead to}

significant distortions within the aromatic rings. Thus, individual Na+_{-C(π) interactions are}

likely weak, but collectively they are sufficiently stabilizing to successfully compete with the more common O-donors of the THF solvent. The Mes-substituted N(4) and N(5) atoms in 4b are significantly pyramidalized (Σ∠(N) = 346.1 and 345.0°), but those substituted with a Ph group are not (av. 357.5°). This distortion is likely due to steric hindrance forcing the Mes group to rotate out of the plane of the ligand backbone, which prevents delocalization of π-electron density from the electron-rich formazanate backbone into the Mes ring. This is also borne out by the large difference between the N-C(Ph) and N-C(Mes) bond lengths (1.379(4)/1.377(3) and 1.435(4)/1.429(4) Å, respectively) that is in agreement with substantial double bond character for the N-C(Ph) groups.

Figure 2.1 Molecular structures of 3b (left), 4a (middle) and {4b}2 (right). Structures are

showing 50% probability ellipsoids. Hydrogen atoms and THF molecules (except for the O atoms bonded to Na) are omitted, and Mes groups in {4b}2 are shown as wireframe for clarity.

2.3.3 Analysis of EPR spectroscopic data of 1-electron reduced products

The solution EPR spectra of compounds 3 are more complex than the nine-line signals usually observed for organic verdazyl radicals,30,31_{and show a multitude of hyperfine interactions.}

Simulation of the spectrum gave a satisfactory fit with inclusion of two pairs of (inequivalent) N atoms as well as the ortho/para phenyl-H and B atoms (Figure 2.2).

(7)

28

Figure 2.2 Experimental and simulated EPR spectrum of 3a (left) and 3b (right). (Simulation parameters: 3a: giso = 2.0057, a(N1,4) = 4.48 G, a(N2,3) = 4.53 G, a(H1) = 1.06 G, a(B) = 0.75 G

and 3b: giso = 2.0040, a(N1,4) = 5.04 G, a(N2,3) = 5.17 G, a(H1) = 1.35 G, a(B) = 0.98 G).

(experimental data taken in THF at 300 K and Simulated by using Easy Spin32,33₎

2.3.4 Characterization of 2-electron reduced products by NMR spectroscopy

In comparison to the neutral compounds 2, the 11_{B NMR resonances in (diamagnetic) dianions}

4 are shifted upfield by ca. 3 ppm, in agreement with more electron-rich compounds. The NMR resonances for the p-CH of the N-Ph groups in dianion 4a are found at δ 5.85 (1_{H) and 109.7}

ppm (13_{C) due to significant charge-delocalization into the aromatic N-substituents of the}

formazanate ligand. For compound 4b, the 1_{H NMR spectrum at room temperature (400 MHz,}

THF-d8 solution) shows exchange broadening of the N-Ph group while the other resonances are

sharp. Upon decreasing the temperature, decoalescence occurs at ca. 15 °C (Figure S2.2) to reveal 5 inequivalent 1_{H environments for the N-Ph group due to restricted rotation around the}

N-C(Ph) bond. In contrast to 4b, the N-Ph resonances in 4a do not show this behavior and exchange-averaged (albeit somewhat broadened) 1_{H NMR resonances are observed down to}

-60 °C. We attribute this difference to the fact that in 4a, resonance delocalization from the electron-rich, reduced formazanate backbone occurs into two aromatic groups (N-Ph), whereas in 4b the perpendicular orientation of the N-Mes group does not allow this and only one N-Ph is involved. As a consequence, in 4b there is more substantial N=C(Ph) double bond character which results in the restricted rotation observed experimentally. NMR lineshape analysis34_for

compound 4b in the temperature range between -30 and + 65 °C allowed determination of the activation parameters for the exchange process as ∆H‡_{= 57.4 ± 1.8 kJ·mol}-1_{and ∆S}‡_{= 1 ± 6}

J·mol-1_·K-1_{(see SI for details). Thus, the barrier to rotation is much higher than measured for}

the N-Ar bond in anilines (∆G‡_{= 25.5 kJ·mol}-1_{for N-methylaniline).}35_{Although it is not as}

3320 3340 3360 3380 3400 3420

Magnetic Field (Gauss)

Experimental Simulated

3320 3340 3360 3380 3400 3420

Magnetic Field (Gauss)

Experimental Simulated N N N N p-tol B Ph Ph 1 2 3 4 H H H H 1 1 1 1 H1 H1 N N N N p-tol B Ph Ph 1 2 3 4 H H 1 1 _H1

(8)

29 high as that in amides (70 – 95 kJ·mol-1_),36,37_{it indicates significant N-C(Ph) π-bonding due to}

the highly reduced nature of the formazanate ligand in 4.

2.3.5 UV-Vis spectroscopy of neutral, 1- and 2-electron reduced formazanate

boron compounds

In order to get further evidence on the ligand-based reduction in [(formazanate)BPh2]0/-1/-2

compounds, UV-Vis absorption spectra of their THF solution have been recorded. Neutral compounds 2a and 2b show absorption in the visible range at 505 nm and 477 nm, respectively, which is due to the π-π* transitions within the formazanate frameworks present in 2a and 2b (Figure S2.1). The absorption of diphenyl compound 2a is blue-shifted compared to its difluoride analogue (LBF2: λmax = 521 nm)13 which is in agreement with a decrease ionic

character in the B-N bonds present in 2a compared to the B-N bonds present in difluoride analogue due to the presence of less electron-withdrawing groups on B (Ph vs F). The absorption of 2b is blue-shifted compared to 2a by 28 nm. This is most likely due to the steric hindrance the NMes group stays out of the plane of ligand core backbone (i.e., NNCNN), and thus, prevents π-conjugation between formazanate backbone and NMes substituent, which makes the length of the π-conjugated system in 2b shorter than that in 2a. UV-Vis spectra of 3a and 3b indicate that each compound has two absorptions in the visible range, one at shorter wavelength (λmax: 488 nm (3a), 485 nm (3b)) and one at longer wavelength ((λmax: 743 nm (3a),

670 nm (3b)) (Figure S2.1). This UV-Vis spectroscopic features of 3a and 3b are consistent with the formation of boron-analogue of a verdazyl radicals30,31_{, which provide further evidence}

for the formation of formazanate-centered radicals11–13_{. The UV-Vis spectra of 4a and 4b show}

absorption bands at 485 nm and 410 nm, respectively, (Figure S2.1). These bands are blue shifted compared to their neutral precursors which indicate further formazanate-centered reduction in 3a and 3b, and suggest the presence of electron-rich trianionic ligand backbones (i.e., L3-_{) in 4a and 4b.}

2.4 Computational studies

Gas phase DFT (B3LYP/6-311+G(d,p)) calculations were performed on the [(formazanate)BPh2]0/-1/-2 compounds in the absence of the cations. The optimized geometries

are in good agreement with the structures determined by crystallography. The N-N bonds in 4acalc and 4bcalc (1.402 – 1.417 Å) are significantly elongated in comparison to the neutral

(9)

30

values. The pyramidalization of the N-Mes moiety observed in the dimer {4b}2 is not

reproduced in 4bcalc, suggesting that this is due to interaction with the bridging Na+ cations.

Analysis of the molecular orbitals shows that the frontier orbitals (LUMO in 2, SOMO in 3 and HOMO in 4) are essentially identical, and are primarily of N-parentage with anti-bonding character between N atoms. An additional contribution comes from the adjacent aromatic rings, which show delocalization on the ortho- and para-positions and have π-bonding character between the N and Ph-Cipso atoms (shown for the HOMO in 4acalc in Figure 2.3, N-C(Ph) =

1.372 Å). This is also reflected by the calculated hyperfine interactions in the radicals 3, with coupling constants of 1.2-1.5 G to the o- and p-H atoms. For compound 4acalc, a scan of the

torsion angle between the N-Ph group and the ligand backbone, followed by transition-state calculations, gave a DFT-computed barrier for rotation around the N-Ph bond of ∆Gǂ_{= 52.9 and}

63.9 kJ·mol-1_{for 4a}_calc_{and 4b}_calc_{, respectively. In the transition state for Ph rotation in 4a}_calc

(Ph/NNCNN dihedral angle of 94°), the N-C bond to the perpendicular Ph group is elongated to 1.418 Å, whereas it is contracted to 1.367 Å for the Ph that is coplanar. The transition state for rotation around the N-C(Ph) bond in 4b has both the N-Ph and N-Mes groups perpendicular, effectively preventing conjugation with the aromatic substituents and localizing the HOMO on the NNCNN backbone.

Figure 2.3 DFT calculated HOMO of 4acalc (left) and transition state of 4acalc (right).

2.5 Conclusions

We have shown that in boron diphenyl compounds, the monoanionic formazanate ligand can be cleanly converted to the corresponding dianionic (radical) and trianionic forms. The highly electron-rich ligand backbone in the latter complexes is stabilized by π-conjugation with the N-Ar substituents, leading to partial double bond character and restricted rotation of the N-C(Ph) bond.

(10)

31

2.6 Experimental section

2.6.1 General considerations

All manipulations were carried out under nitrogen or argon atmosphere using standard glovebox, Schlenk, and vacuum-line techniques. Toluene and hexane (Aldrich, anhydrous, 99.8%) were passed over columns of Al2O3 (Fluka), BASF R3-11-supported Cu oxygen scavenger, and

molecular sieves (Aldrich, 4 Å). THF (Aldrich, anhydrous, 99.8%) was dried by percolation over columns of Al2O3 (Fluka). For synthesis of the very sensitive compounds 4, THF and

hexane were additionally dried on Na/K alloy and subsequently vacuum transferred and stored under nitrogen. All solvents were degassed prior to use and stored under nitrogen. C6D6

(Aldrich) and d8-THF (Sigma-Aldrich) were vacuum transferred from Na/K alloy and stored under nitrogen. The compound {[MesN2]+[BF4]-}38 and ligands PhNNC(p-tolyl)NNPh (1a)38

& PhNNC(p-tolyl)NNMes (1b)39_{were synthesized according to published procedures. NMR}

spectra were recorded on Varian Mercury 400 or Inova 500 spectrometers. The 1_{H and}13_{C NMR}

spectra were referenced internally using the residual solvent resonances and reported in ppm relative to TMS (0 ppm); J is reported in Hz. Assignments of NMR resonances was aided by COSY, NOESY, HSQC and HMBC experiments using standard pulse sequences. Elemental analyses were performed at the Microanalytical Department of the University of Groningen or Kolbe Microanalytical Laboratory (Mülheim an der Ruhr, Germany). UV-Vis spectra were recorded in THF solution (~ 10-4_{M) using a Perkin Elmer Lambda 900 in a quartz cell that was}

sealed under N2 atmosphere. EPR spectra were recorded on a Bruker EMXplus spectrometer at

300K in THF solution. Cyclic voltammetry (CV) was performed using a three-electrode configuration comprising a Pt wire counter electrode, a Ag wire pseudo reference electrode and a Pt disk working electrode (CHI102, CH Instruments, diameter = 2 mm). The Pt working electrode was polished before each experiment using an alumina slurry (0.05 μm), rinsed (thrice) with distilled water and acetone. The electrodes were then dried in an oven at 75 °C for at least one hour to remove any residual traces of water. The CV data were referenced by measuring the redox potential of ferrocene, which was added as a THF solution at the end of the experiment. In all cases, there is no indication that addition of ferrocene influences the electrochemical behavior. All electrochemical measurements were performed at ambient temperatures in a nitrogen glovebox with the compounds dissolved in THF containing 0.1 M [n_Bu₄_N][PF₆_{] as the supporting electrolyte. The electrochemical data were measured using an}

(11)

32

2.6.2 Compounds synthesis and characterization

Synthesis of (PhNNC(p-tol)NNPh)BPh2 (2a). 3-p-tolyl-1,5-diphenyl-formazan 1a (1.005 g,

3.2 mmol) and triphenyl borane (1.55 g, 6.4 mmol) were weighed out inside a glovebox, added into a 250 mL Schlenk flask and subsequently dissolved by addition of 110 mL of dry toluene. The mixture was heated to 125 ◦_{C under an inert atmosphere for 3 days. After this the solvent}

was evaporated. NMR analysis of the crude product (C6D6) indicated full conversion of the

formazan starting material and the presence of minor impurities. Purification was achieved by column chromatography (Al2O3; hexane:DCM = 7:3). The solvent was removed under reduced

pressure and the solid product was recrystallized by cooling a hot hexane solution to -30 °C. After standing at -30 °C for 1 day, crystals had separated which were collected by filtration. Subsequent drying gave 2a as dark pink crystalline material (970 mg, 2.03 mmol, 63 %). 1_H

NMR (400 MHz, C6D6, 25◦C) δ 8.02 (d, J = 8.1 Hz, 2H, p-tolyl-o-H), 7.40 – 7.29 (m, 8H,

(B-Ph & N-(B-Ph) m-H), 7.09–7.02 (m, 6H, B-(B-Ph o-H & p-H), 6.95 (d, J = 8.1 Hz, 2H, p-tolyl-m-H), 6.86 – 6.77 (m, 6H, N-Ph o-H & p-H), 2.02 (s, 3H, p-tolyl-CH3). 11B NMR (128.3 MHz, C6D6,

25 °C) δ 1.74 (s ). 13_{C NMR (100 MHz, C}₆_D₆_{, 25}◦_{C) δ 154.37 (NCN), 146.91 (N-C(ipso)-Ph),}

143.80 (B-C(ipso)-Ph), 139.03 (p-tolyl CH3), 135.30 (B-Ph m-CH), 131.46

(p-tolyl-C(ipso)-NCN), 129.60 (p-tolyl m-CH), 128.44 (N-Ph CH), 127.94 (N-Ph p-CH), 127.61 (B-Ph o-CH), 127.33 (B-Ph p-o-CH), 126.78 (N-Ph m-o-CH), 125.60 (p-tolyl o-o-CH), 21.23 (p-tol CH3).

Anal. Calcd for C32H27BN4: C, 80.34; H, 5.69; N, 11.71. Found: C, 80.05; H, 5.71; N, 11.59.

Synthesis of [(PhNNC(p-tol)NNPh)BPh2][Cp*2Co].(2THF) (3a). Compound 2a (151 mg,

0.316 mmol) and decamethylcobaltocene (Cp*2Co) (107 mg, 0.325 mmol) were mixed in 2ml

dry THF in a vial and the reaction mixture was kept stirring for 12 h, during which time the mixture turned green. Filtration and addition of hexane (4 mL) precipitated a green crystalline material. The mixture was subsequently cooled to -30 °C to complete precipitation of the product. The intense green color crystals were washed with pentane (3 x 2 mL) and dried to give 3a as green crystals (201 mg, 0.228 mmol, 72 %).

Based on the X-ray analysis, there are 2 THF solvate molecules present per boron fragment, one of which is somewhat disordered. Elemental analysis data for crystalline 3a are in better agreement with the presence of 1 THF molecule, which suggests that the disordered THF may have only partial occupancy, or is partially removed upon drying the crystals. For comparison, we report here the calculated analysis data with 1 and 2 THF molecules present. The found analysis values are consistent with a THF/B ration of somewhere in between 1 and 2. With 1

(12)

33 THF, C56H65BCoN4O: C, 76.44; H, 7.45; N, 6.37. With 2 THF, C60H73BCoN4O2: C, 75.70; H,

7.73; N, 5.89. Found: C, 76.51; H, 7.98; N, 5.88.

[(PhNNC(p-tol)NNPh)BPh2][Na2(THF)6].THF (4a). To a solution of naphthalene (73.05 mg,

0.57 mmol) in 4 mL of THF was added sodium metal (19.55 mg, 0.85 mmol). After stirring for 12 h, the mixture was filtered and the green solution was added to a vial containing 2a (134 mg, 0.28 mmol). The reaction mixture was stirred for 12 h, during which it changed color from dark pink to green and finally orange. A layer of hexane (3 mL) was added onto the THF solution, and after slow diffusion of the two layers the mixture was stored at -30 °C for 2 more days. The resulting orange crystals were washed with pentane (3 x 2 mL) and dried to give compound 4a as a highly air-sensitive solid (150 mg, 0.157 mmol, 56 %). 1_{H NMR (400 MHz, THF-d}₈_{, 25}◦_C)

δ 8.17 (d, J = 7.9 Hz, 2H, p-tol o-H), 7.62 (d, J = 7.1 Hz, 4H, BPh o-H), 7.03 (d, J = 7.8 Hz, 2H, p-tol m-H), 6.95 (q, J = 7.4 Hz, 8H, BPh & NPh m-H), 6.84 (t, J = 7.1 Hz, 2H, BPh p-H), 6.55 (t, J = 7.7 Hz, 4H, NPh o-H), 5.85 (t, J = 6.9 Hz, 2H, NPh p-H), 3.62 (t, J = 6.4 Hz, 14H, THF), 2.30 (s, 3H, p-tol CH3), 1.82 – 1.70 (m, 14H, THF). 11B NMR (128.3 MHz, d8-THF, 25

°C) δ -0.99 (s). 13_{C NMR (125 MHz, THF-d}₈_{, 25}◦_{C) δ 153.07 (N-C(ipso) Ph), 152.71 (NCN),}

142.94 (p-tol CCH3), 136.11 (B-C(ipso) Ph), 135.46 (p-tol C(ipso)-NCN), 135.08 (BPh o-CH),

128.16 (p-tol m-CH), 128.06 (NPh o-CH), 127.65 (p-tol o-CH) , 127.18 (BPh m-CH), 124.00 (BPh p-CH), 114.11 (NPh m-CH), 109.69 (NPh p-CH), 68.39 (THF), 26.56 (THF), 21.46 (p-tol CH3). According to X-ray analysis, the crystals contain 6 molecules of THF that are bound

to Na+_{cations, and one THF solvate molecule. Drying of the crystals changed their lustrous}

appearance, indicating that (partial) loss of the THF solvate may be occurring. Due to its highly air-sensitive nature combined with the facile loss of solvent, all our elemental analysis measurements gave consistently low values, especially for C and H analyses.

Synthesis of (PhNNC(p-tol)NNMes)BPh2 (2b). 3-p-tolyl-1-mesityl-5-phenyl-formazan 1b

(400 mg, 1.12 mmol) and triphenyl borane (406.8 mg, 1.68 mmol) were weighed out inside a glove box and added into a 250 mL Schlenk flask. After dissolution n 60 mL of dry toluene, the mixture was heated and stirred at 120 ◦_{C for 2 days. After this the solvent was evaporated. The}

crude product was purified by column chromatography (SiO2; hexane:DCM = 4:1) and the

solvent was evaporated on the rotavap. The product was subsequently crystallized by cooling a hot hexane solution to room temperature and then to -30 °C overnight. The precipitate was collected by filtration to give 2b as intense pink crystalline product (430 mg, 0.826 mmol, 74 %). 1_{H NMR (400 MHz, C}₆_D₆_{, 25}◦_{C) δ 7.96 (d, J = 8.0 Hz, 2H, p-tol o-H), 7.53 – 7.48 (m, 2H,}

(13)

34

2H, p-tol m-H), 6.88 (overlapped d, 2H, NPh o-H), 6.87 (overlapped m, 1H, NPh p-H), 6.40 (s, 2H, Mes m-H), 2.00 (s, 3H, p-tol CH3), 1.99 (s, 6H, Mes o-CH3), 1.89 (s, 3H, Mes p-CH3). 11B

NMR (128.3 MHz, C6D6, 25 °C) δ 2.27 (s). 13C NMR (100 MHz, C6D6, 25◦C) δ 154.66 (NCN),

145.54 (N-C(ipso) Ph), 143.13 (N-C(ipso) Mes), 142.1 (B-C(ipso) Ph), 139.18 (p-tol CCH3),

138.30 (Mes-C(ipso) p-CH3), 136.40 (Mes-C(ipso) o-CH3), 136.02 (BPh m-CH), 131.19

(NCN-C-tol), 129.63 (Mes-CH), 129.59 (tol o-CH), 128.41 (NPh o-CH), 128.19 (NPh p-CH), 127.19 (NPh m-p-CH), 127.08 (BPh o-p-CH), 127.02 (BPh p-p-CH), 125.86 (p-tol m-p-CH), 21.22 (p-tol CH3), 20.77 (Mes p-CH3), 19.74 (Mes o-CH3). Despite repeated attempts, samples

of 2b consistently gave low analysis values for C content, while H and N analyses gave satisfactory agreement with calculated values. Anal. Calcd for C35H33BN4: C, 80.77; H, 6.39;

N, 10.76. Found: C, 79.81; H, 6.39; N, 10.61.

Synthesis of [(PhNNC(p-tol)NNMes)BPh2][Cp*2Co].THF (3b). Compound 2b (50 mg, 0.096

mmol) and decamethylcobaltocene (31.64 mg, 0.096 mmol) were mixed in a vial and stirred with 1 mL of dry THF. After 12 h, the green reaction mixture was filtered and a layer of hexane was added on top of the THF solution. Slow interdiffusion of the layers resulted in the precipitation of green crystals. After removal of the supernatant, the crystals were washed with pentane (3 x 1 mL) and dried to give 3b as green crystalline material (62 mg, 0.073 mmol, 76 %). Anal. Calcd for C59H71BN4CoO: C, 76.86; H, 7.76; N, 6.08. Found: C, 76.79; H, 7.81; N,

5.93.

Synthesis of [(PhNNC(p-tol)NNMes)BPh2]2[Na4(THF)7] (4b). A solution of NaC10H8 was

prepared as described for 4a (naphthalene: 24.6 mg, 0.192 mmol; Na: 6.6 mg, 0.288 mmol; in 1 mL of THF). This solution was added to compound 2b (50 mg, 0.096 mmol), which caused the color of the solution to change from pink to green and finally orange. After stirring for 12 h, hexane (2 mL) was added on top of the THF solution which, after standing for 2 days at room temperature, caused the precipitation of orange crystals. The crystals were washed with pentane (3 x 1 mL) and dried to give 4b as a highly air-sensitive solid (53 mg, 0.064 mmol, 67 %). 1_H

NMR (500 MHz, THF-d8, -30 ◦C) δ 7.70 (d, J = 7.3 Hz, 2H, p-tol o-H), 7.31 (br, 4H, B-Ph,

m-H), 7.17 (d, J = 8.0 Hz, 1H, N-Ph o-m-H), 6.88 (d, J = 7.4 Hz, 2H, p-tol m-m-H), 6.77-6.72 (bd, J = 18.2 Hz, 7H, overlapped Ph m-H (1H) + B-Ph o/p-H (6H)), 6.29 (s, 2H, N-Mes m-H), 6.01 (t, J = 7.6 Hz, 1H, N-Ph m-H), 5.76 (d, J = 8.6 Hz, 1H, N-Ph o-H), 5.57 (t, J = 6.6 Hz, 1H, N-Ph p-H), 3.61 (m, THF), 2.21 (m, 3H, p-tol CH3), 1.99 (s, 3H, Mes p-CH3), 1.94 (s, 6H, Mes

o-CH3), 1.78 (m, THF). 11B NMR (128.3 MHz, C6D6, 25 °C) δ -0.97 (s). 13C NMR (125 MHz,

(14)

35 (NCN-C(ipso)-p-tol), 136.35 (Mes-C(ipso)-o-CH3), 136.20 (BPh m-CH), 134.29 (p-tol

C(ipso)-p-CH3), 129.18 (NPh m-CH), 128.87 CH), 127.93 (p-tol m-CH), 126.60

(Mes-C(ipso)-p-CH3), 126.56 (p-tol o-CH3), 126.41 (NPh m-CH), 125.53 (BPh o-CH), 123.12 (BPh

p-CH), 117.59 (NPh o-CH), 109.06 (NPh o-CH), 106.15 (NPh p-CH), 68.29 (THF), 26.45 (THF), 21.38 (Mes o-CH3), 21.34 (p-tol CH3), 21.12 (Mes p-CH3).

According to X-ray analysis, the crystals contain 7 molecules of THF that are bound to 4 Na+

cations, and disordered electron density that is likely due to the present of one (or more) solvent molecules (THF and/or hexane). Drying of the crystals changed their lustrous appearance, indicating that (partial) loss of the solvent molecules may be occurring. Due to this, and its high sensitivity, a satisfactory elemental analysis could not be obtained despite repeated attempts.

2.6.3 X-ray crystallography

Suitable crystals of compounds 2a, 3a, 3b, 4a and 4b were mounted on top of a cryoloop and transferred into the cold (100 K) nitrogen stream of a Bruker D8 Venture diffractometer. Data collection and reduction was done using the Bruker software suite APEX240_{. Data collection}

was carried out at 100 K using Mo radiation (0.71073 Å). The final unit cell was obtained from the xyz centroids of 9867 (2a), 9792 (3a), 9798 (3b), 9668 (4a) or 9823 (4b) reflections after integration. A multiscan absorption correction was applied, based on the intensities of symmetry-related reflections measured at different angular settings (SADABS)40_._{The structures}

were solved by direct methods using SHELXS, and refinement of the structure was performed using SHLELXL41_{. For 3a, refinement indicated the presence of two THF solvate molecules, of}

which one was somewhat disordered and/or partially occupied. Removal of its contribution using the PLATON/SQUEEZE routine42_{allowed convergence of the refinement. Refinement}

of 4b was frustrated by a disorder problem: several THF molecules bound to the Na cations showed high anisotropic displacement parameters. The THF-carbon atoms that were most affected by this disorder were split in two components for which site-occupancy factors were refined. In addition, large residual peaks were observed in the difference Fourier map, which is likely due to disordered solvate molecules. The contribution from this region was removed using the PLATON/SQUEEZE routine42_{. In the final stages of refinement, some atoms showed}

unrealistic displacement parameters, and one atom (C94A) refined to non-positive definite values. Ultimately, SIMU and DELU instructions were applied. The hydrogen atoms were generated by geometrical considerations, constrained to idealized geometries and allowed to ride on their carrier atoms with an isotropic displacement parameter related to the equivalent

(15)

36

displacement parameter of their carrier atoms. Crystal data and details on data collection and refinement are presented in Table 2.1.

Table 2.1 Crystallographic data for 2a, 3a, 3b, 4a and {4b}2

2a 3a 3b 4a {4b}2 chem formula C32 H27 B N4 C56 H65 B Co N4 O C59 H71 B Co N4 O C60 H83 B N4 Na2 O7 C98 H122 B2 N8 Na4 O7 Mr 478.38 879.86 921.93 1029.09 1637.61

cryst syst monoclinic monoclinic monoclinic triclinic triclinic color, habit red, block green, platelet green, platelet orange, block orange, block size (mm) 0.32 x 0.18 x 0.08 0.22 x 0.21 x 0.04 0.30 x 0.20 x 0.07 0.39 x 0.21 x 0.19 0.47 x 0.40 x 0.33 space group P21/c P21/c C2/c P-1 P-1 a (Å) 9.5375(3) 13.3417(6) 42.008(4) 11.0324(4) 13.2345(9) b (Å) 15.5183(5) 14.8614(7) 11.3469(10) 12.8217(5) 19.0244(14) c (Å) 17.4792(6) 25.5659(12) 21.4239(18) 20.6913(8) 22.8548(15) α (°) 78.4851(12) 88.975(2) β (°) 99.2996(12) 92.1498(16) 101.719(3) 84.7390(17) 78.902(2) γ (°) 79.0342(14) 81.576(2) V (Å3₎ _2553.02(14) _5065.5(4) _9999.2(15) _2811.20(19) _5585.5(7) Z 4 4 8 2 2 ρcalc, g.cm-3 1.245 1.154 1.225 1.216 0.974 Radiation [Å] 0.71073 0.71073 0.71073 0.71073 0.71073 µ(Mo Kα), mm-1 0.074 0.380 0.388 0.092 0.074 F(000) 1008 1876 3944 1108 1752 temp (K) 100(2) 100(2) 100(2) 100(2) 100(2) θ range (°) 2.879 - 27.148 2.855 - 25.027 2.834 - 27.195 2.712 - 26.372 2.877 - 24.814

(16)

37 data collected (h,k,l) -12:12; -15:19; -22:22 -15:15; -17:17; -30:30 -53:53; -14:14; -27:27 -13:13; -16:16; -25:25 -15:15; -22:22; -26:26 no. of rflns collected 31641 74035 108255 104408 86821 no. of indpndt reflns 5633 8925 11087 11468 19067 observed reflns Fo ≥ 2.0 σ (Fo) 4698 6889 9613 9320 13763 R(F) (%) 3.75 6.19 6.56 7.21 6.98 wR(F2_{) (%)} _9.67 _13.57 _13.04 _15.41 _19.95 GooF 1.036 1.103 1.187 1.146 1.061 weighting a,b 0.0416, 1.0584 0.0281, 15.1406 0.0230, 47.2807 0.0373, 4.8816 0.0869, 7.4986 params refined 335 579 612 678 1120 min, max resid dens -0.212, 0.305 -0.380, 0.544 -0.807, 0.637 -0.487, 0.559 -0.487, 0.559

Figure 2.4 Molecular structure of 2a (left) and 3a (right) showing 50% probability ellipsoids, hydrogen atoms and THF solvate molecule (for 3a) omitted for clarity.

(17)

38

2.7 Supplementary information

2.7.1 UV-Vis spectra

Figure S2.1 UV-Vis spectra of 2a, 3a and 4a (left) and 2b, 3b and 4b (right) in THF solution.

2.7.2 Determination of exchange kinetics in compound 4b

1_{H NMR data for compound 4b were collected in the temperature range 243 – 338 K (Figure}

S2.2). The Varian NMR data files were converted to the gNMR34_{file format using the gCVT}

tool included in the gNMR installation. The chemical shifts of the peaks of interest (those of the N-Ph group) were taken from the experimental spectrum and exchange between pairs (the two ortho-H and meta-H) was modeled; the signal for the para-H was included without exchange. The latter peak was used to estimate the linewidth in the absence of chemical exchange (due to relaxation, inhomogeneity of the magnetic field etc.). Additional line broadening due to chemical exchange was then included, and the agreement between experimental and simulated spectrum was inspected visually. Due to the presence of additional peaks in the region of interest, attempts to perform least-squares fitting of the line shapes were unsuccessful. An estimate of the error in the exchange rate constants was made visually by running simulations with different rate constants and evaluating in which range a satisfactory fit was still obtained. A comparison between experimental spectra and those with ‘best’ fit parameters are shown in Figure S2.3.

The rate constants thus obtained were used for constructing an Eyring plot of Ln(k/T) vs. 1/T. The estimated errors were taken into account by giving each data point a weight that was proportional to 1/(σ(k)2_{). Fitting was performed using Wolfram Mathematica 10,}43_and

activation parameters are determined using standard procedures from the slope and intercept.

-0.05 0.45 0.95 1.45 1.95 2.45 2.95 3.45 3.95 200 400 600 800 1000 ε( L. mo l -1.cm -1) x 1 00 00 Wavelength (nm) 2a 3a 4a -0.05 0.95 1.95 2.95 3.95 4.95 200 400 600 800 1000 ε( L. mo l -1.cm -1) x 1 00 00 Wavelength (nm) 2b 3b 4b

(18)

39 Figure S2.2 1_{H NMR of compound 4b at various temperatures in THF-d}₈

243 K 288 K 278 K 298 K 308 K 328 K 318 K 263 K 338 K

(19)

40

Figure S2.3. Comparison of experimental and simulated 1_{H-NMR spectra of 4b (for each}

temperature, top: experimental spectrum, bottom: simulated). Rate constants used for the simulation are shown for each spectrum, including an estimate of the error.

7.000 7.000 6.750 6.750 6.500 6.500 6.250 6.250 6.000 6.000 5.750 5.750 7.000 7.000 6.750 6.750 6.500 6.500 6.250 6.250 6.000 6.000 5.750 5.750 p 7.000 7.000 6.750 6.750 6.500 6.500 6.250 6.250 6.000 6.000 5.750 5.750 p 7.000 7.000 6.750 6.750 6.500 6.500 6.250 6.250 6.000 6.000 5.750 5.750 p 7.000 7.000 6.750 6.750 6.500 6.500 6.250 6.250 6.000 6.000 5.750 5.750 p 7.000 7.000 6.750 6.750 6.500 6.500 6.250 6.250 6.000 6.000 5.750 5.750 p 7.000 7.000 6.750 6.750 6.500 6.500 6.250 6.250 6.000 6.000 5.750 5.750 p 7.000 7.000 6.750 6.750 6.500 6.500 6.250 6.250 6.000 6.000 5.750 5.750 7.000 7.000 6.750 6.750 6.500 6.500 6.250 6.250 6.000 6.000 5.750 5.750

243 K: estimated rate constant k = 5 ± 4 s-1

298 K: estimated rate constant k = 600 ± 100 s-1 _{308 K: estimated rate constant k = 800 ± 200 s}-1

338 K: estimated rate constant k = 12000 ±

2000 s-1

(20)

41

Figure S2.4 (a) Eyring plot for the calculation of activation parameters for rotation around the N-C(Ph) bond in compound 4b.

2.7.3 Computational studies

Calculations were performed with the Gaussian09 program44_{using density functional theory}

(DFT) in the gas phase. Geometries were fully optimised starting from the X-ray structures using the B3LYP exchange-correlation functional with the 6-311+G(d,p) basis set. Optimizations were performed without (symmetry) constraints, and the resulting structures were confirmed to be minima on the potential energy surface by frequency calculations (number of imaginary frequencies = 0). The stationary point found for 2acalc closely resembles the

structure determined by X-ray crystallography. For the singly reduced compounds 3a and 3b, geometry optimizations were performed on the isolated anions [L1BPh2]-. (3acalc) and [L2BPh2] -._(3b_calc_{) and the converged optimized geometries were in good agreement with the structures}

determined by X-ray crystallography. For the doubly reduced compounds 4, geometry optimizations were performed on the isolated (monomeric) dianions [L1BPh2]2- (4acalc) and

[L2BPh2]2- (4bcalc).

To explore the chemical exchange process of 4b due to the rotation of N-Ph substituent around the N-C(Ph) bond we scanned the dihedral angle between the two adjacent N atoms in the ligand and the ipso- and ortho-carbons of the Ph group in steps of 10 degree (all other degrees of freedom were optimized). Starting from the highest energy points along the scan coordinate, the transition states were optimized and confirmed by frequency analysis (number of imaginary frequencies for 4acalc-ts and 4bcalc-ts =1).

(21)

42

Figure S2.5 HOMOs of 4bcalc (left) and the transition state for N-Ph rotation of 4bcalc (4bcalc-ts,

right)

Table S2.1 Comparison of selected bond lengths values (Å) between X-ray measured and DFT calculated for 2a, 3a & 4a

Bonds Bond lengths

2a 2acalc 3a 3acalc 4a 4acalc

N1-N2 1.306(1) 1.298 1.371(4) 1.352 1.428(3) 1.402 N2-C7 1.344(1) 1.343 1.345(5) 1.334 1.325(3) 1.328 C7-N3 1.347(1) 1.343 1.338(5) 1.334 1.326(3) 1.327 N3-N4 1.309(1) 1.298 1.370(4) 1.352 1.433(3) 1.402 N1-B1 1.599(2) 1.610 1.569(5) 1.594 1.578(3) 1.590 N4-B1 1.595(1) 1.609 1.567(5) 1.594 1.583(3) 1.590 N1-C1 1.434(2) 1.430 1.400(4) 1.404 1.379(3) 1.372 N4-C15 1.433(1) 1.430 1.403(4) 1.404 1.384(3) 1.372 B1-C21 1.618(2) 1.624 1.635(6) 1.640 1.637(4) 1.651 B1-C27 1.604(2) 1.616 1.613(5) 1.633 1.626(3) 1.645

(22)

43 Table S2.2 Comparison of selected bond lengths values (Å) between X-ray measured and DFT calculated for 2b, 3b & (4b)2

Bonds Bond lengths Bonds Bond lengths

2bcalc 3b 3bcalc (4b)2 4bcalc

N1-N2 1.301 1.36(3) 1.356 N1-N2/N5-N6 1.431(3)/1.428(3) 1.402 N2-C7 1.342 1.332(3) 1.334 N2-C7/N6-C42 1.319(4)/1.319(3) 1.330 C7-N3 1.347 1.342(3) 1.341 C7-N3/C42-N7 1.337(4)/1.337(4) 1.330 N3-N4 1.298 1.354(3) 1.354 N3-N4/N7-N8 1.469(3)/1.470(3) 1.417 N1-B1 1.615 1.589(4) 1.600 N1-B1/N5-B2 1.569(4)/1.577(4) 1.608 N4-B1 1.606 1.587(3) 1.586 N4-B1/N8-B2 1.583(4)/1.580(4) 1.568 N1-C1 1.431 1.407(3) 1.400 N1-C1/N5-C36 1.378(4)/1.376(3) 1.368 N4-C15 1.449 1.428(3) 1.431 N4-C15/N8-C50 1.436(4)/1.429(4) 1.400 B1-C24 1.631 1.639(4) 1.648 B1-C24/B2-C59 1.637(4)/1.632(4) 1.643 B1-C30 1.617 1.618(4) 1.633 B1-C30/B2-C65 1.645(4)/1.636(4) 1.656

(23)

44

2.8 References

(1) Lyaskovskyy, V.; De Bruin, B. ACS Catal. 2012, 2 (2), 270–279.

(2) Luca, O. R.; Crabtree, R. H. Chem. Soc. Rev. 2013, 42 (4), 1440–1459.

(3) Demir, S.; Jeon, I. R.; Long, J. R.; Harris, T. D. Coord. Chem. Rev. 2015, 289–290 (1), 149–176.

(4) Avent, A. G.; Khvostov, A. V; Hitchcock, P. B.; Lappert, M. F. Chem. Commun. 2002, 415 (13), 1410–

1411.

(5) Eisenstein, O.; Hitchcock, P. B.; Khvostov, A. V; Lappert, M. F.; Maron, L.; Perrin, L.; Protchenko, A. V.

J. Am. Chem. Soc. 2003, 125 (36), 10790–10791.

(6) Avent, A. G.; Hitchcock, P. B.; Khvostov, A. V.; Lappert, M. F.; Protchenko, A. V. Dalt. Trans. 2004, No.

15, 2272–2280.

(7) Ibrahim, S. K.; Khvostov, A. V.; Lappert, M. F.; Maron, L.; Perrin, L.; Pickett, C. J.; Protchenko, A. V.

Dalt. Trans. 2006, No. 21, 2591–2596.

(8) Coles, M. P.; Hitchcock, P. B.; Khvostov, A. V.; Lappert, M. F.; Protchenko, A. V. Dalton Trans. 2010, 39

(28), 6426.

(9) S. Itoh and Y. Morimoto, in Chemical Science of π-Electron Systems, Eds. T. Akasaka, A. Osuka, S.

Fukuzumi, H. Kandori and Y. Aso, Springer Japan, 2015, Ch. 42, Pp. 715-730.

(10) Camp, C.; Arnold, J. Dalton Trans. 2016, 45 (37), 14462–14498.

(11) Gilroy, J. B.; Ferguson, M. J.; McDonald, R.; Patrick, B. O.; Hicks, R. G. Chem. Commun. 2007, 412 (2),

126–128.

(12) Chang, M.-C.; Dann, T.; Day, D. P.; Lutz, M.; Wildgoose, G. G.; Otten, E. Angew. Chem. Int. Ed 2014, 53

(16), 4118–4122.

(13) Chang, M.-C.; Otten, E. Chem. Commun. 2014, 50 (56), 7431–7433.

(14) Chang, M. C.; Chantzis, A.; Jacquemin, D.; Otten, E. Dalton Trans. 2016, 45 (23), 9477–9484.

(15) Travieso-Puente, R.; Broekman, J. O. P.; Chang, M. C.; Demeshko, S.; Meyer, F.; Otten, E. J. Am. Chem.

Soc. 2016, 138 (17), 5503–5506.

(16) Hong, S.; Hiii, L. M. R.; Gupta, A. K.; Naab, B. D.; Gllroy, J. B.; Hicks, R. G.; Cramer, C. J.; Tolman, W. B. Inorg. Chem. 2009, 48 (10), 4514–4523.

(17) Barbon, S. M.; Price, J. T.; Reinkeluers, P. A.; Gilroy, J. B. Inorg. Chem. 2014, 53 (19), 10585–10593.

(18) Barbon, S. M.; Reinkeluers, P. A.; Price, J. T.; Staroverov, V. N.; Gilroy, J. B. Chem. - A Eur. J. 2014, 20

(36), 11340–11344.

(19) Hesari, M.; Barbon, S. M.; Staroverov, V. N.; Ding, Z.; Gilroy, J. B. Chem. Commun. 2015, 51 (18), 3766–

3769.

(20) Maar, R. R.; Barbon, S. M.; Sharma, N.; Groom, H.; Luyt, L. G.; Gilroy, J. B. Chem. - A Eur. J. 2015, 21

(44), 15589–15599.

(21) Kabir, E.; Wu, C.-H.; Wu, J. I. C.; Teets, T. S. Inorg. Chem. 2016, 55 (2), 956–963.

(22) Schorn, W.; Grosse-Hagenbrock, D.; Oelkers, B.; Sundermeyer, J. Dalton Trans. 2016, 45 (3), 1201–1207.

(23) Chang, M.-C.; Otten, E. Inorg. Chem. 2015, 54 (17), 8656–8664.

(24) Gokel, G. W.; Barbour, L. J.; Ferdani, R.; Hu, J. Acc. Chem. Res. 2002, 35 (10), 878–886.

(25) Salonen, L. M.; Ellermann, M.; Diederich, F. Angew. Chemie - Int. Ed. 2011, 50 (21), 4808–4842.

(24)

45 (27) Fukin, G. K.; Lindeman, S. V.; Kochi, J. K. J. Am. Chem. Soc. 2002, 124 (28), 8329–8336.

(28) Torvisco, A.; Ruhlandt-Senge, K. Inorg. Chem. 2011, 50 (24), 12223–12240.

(29) McWilliams, S. F.; Rodgers, K. R.; Lukat-Rodgers, G.; Mercado, B. Q.; Grubel, K.; Holland, P. L. Inorg.

Chem. 2016.

(30) R. G. Hicks, John Wiley & Sons, Ltd, 2010, pp. 245-279. No Title.

(31) Matuschek, D.; Eusterwiemann, S.; Stegemann, L.; Doerenkamp, C.; Wibbeling, B.; Daniliuc, C. G.; Doltsinis, N. L.; Strassert, C. A.; Eckert, H.; Studer, A. Chem. Sci. 2015, 6 (8), 4712–4716.

(32) Stoll, S.; Schweiger, A. EasySpin: Simulating Cw ESR Spectra; 2007; Vol. 27. (33) Stoll, S.; Schweiger, A. J. Magn. Reson. 2006, 178 (1), 42–55.

(34) P. H. M. Budzelaar, gNMR v. 5. 0. 6. https://home. cc. umanitoba. ca/~budzelaa/gNMR/gNMR. html. (35) Anet, F. A. L.; Mehran, G. J. Am. Chem. Soc. 1979, 101 (23), 6857–6860.

(36) Stewart, W. E.; Siddall, T. H. Chem. Rev. 1970, 70 (5), 517–551.

(37) Dugave, C.; Demange, L. Chem. Rev. 2003, 103 (7), 2475–2532.

(38) Doyle, M. P.; Bryker, W. J. J. Org. Chem. 1979, 44 (9), 1572–1574.

(39) Chang, M.-C.; Roewen, P.; Travieso-Puente, R.; Lutz, M.; Otten, E. Inorg. Chem. 2015, 54 (1), 379–388.

(40) Bruker. APEX2 (v2012.4-3), SAINT (Version 8.18C) and SADABS (Version 2012/1). Bruker AXS Inc., Madison, Wisconsin, U. 2012.

(41) Sheldrick, G. M. Acta Crystallogr. Sect. A Found. Adv. 2015, 71 (1), 3–8.

(42) Spek, A. L. J. Appl. Crystallogr. 2003, 36 (1), 7–13.

(43) Wolfram Research, Inc., Mathematica 10.3, Champaign, I. (2016).

(44) Gaussian 09, Revision D.01, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, Ö. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, and D. J. Fox, Gaussian, Inc., Wallingford CT, 2009.

(45) GaussView, Version 5, Dennington, Roy; Keith, Todd; Millam, John. Semichem Inc., Shawnee Mission, KS, 2009

(25)