Revealing redox isomerism in trichromium imides by anomalous diffraction

In polynuclear biological active sites, multiple electrons are needed for turnover, and the distribution of these electrons among the metal sites is affected by the structure of the active site. However, the study of the interplay between structure and redox distribution is difficult not only in biological systems but also in synthetic polynuclear clusters since most redox changes produce only one thermodynamically stable product. Here, the unusual chemistry of a sterically hindered trichromium complex allowed us to probe the relationship between structural and redox isomerism. Two structurally isomeric trichromium imides were isolated: asymmetric terminal imide (tbsL)Cr3(NDipp) and symmetric, μ3-bridging imide (tbsL)Cr3(μ3–NBn) ((tbsL)6− = (1,3,5-C6H9(NC6H4-o-NSitBuMe2)3)6−). Along with the homovalent isocyanide adduct (tbsL)Cr3(CNBn) and the bisimide (tbsL)Cr3(μ3–NPh)(NPh), both imide isomers were examined by multiple-wavelength anomalous diffraction (MAD) to determine the redox load distribution by the free refinement of atomic scattering factors. Despite their compositional similarities, the bridging imide shows uniform oxidation of all three Cr sites while the terminal imide shows oxidation at only two Cr sites. Further oxidation from the bridging imide to the bisimide is only borne at the Cr site bound to the second, terminal imido fragment. Thus, depending on the structural motifs present in each [Cr3] complex, MAD revealed complete localization of oxidation, partial localization, and complete delocalization, all supported by the same hexadentate ligand scaffold.


General considerations
All manipulations involving metal complexes were carried out using standard Schlenk or glovebox techniques under a dinitrogen atmosphere, unless otherwise noted. All glassware was oven-dried for a minimum of 10 h and cooled in an evacuated chamber prior to use in the drybox. was prepared by diazotization of the corresponding aniline and subsequent treatment with NaN3.
All organic azides were dissolved in hexanes and stored over 4 Å molecular sieves prior to use, at which point the solution was decanted and the hexanes was evaporated. All other reagents were purchased from commercial vendors and used without further purification.

Electrochemical measurements
Cyclic voltammetry was performed with a CHI660d potentiostat using a three-electrode cell with a glassy carbon working electrode and a platinum wire as the counter electrode. All measurements were conducted using a Ag/AgCl pseudoreference consisting of a silver wire immersed in 0.

Magnetic susceptibility measurements
Magnetic data for 1-5 were collected using a Quantum Design MPMS-XL Evercool SQUID Magnetometer. A general procedure for sample preparation is as follows: microcrystalline material was dried thoroughly in vacuo overnight and then crushed to a fine powder in an agate mortar and pestle. This crushed powder was then immobilized within a gelatin capsule size #4 by adding melted eicosane at 50 -60 °C. The gelatin capsule was then inserted into a plastic straw.
Samples were prepared under a dinitrogen atmosphere. Magnetization data at 100 K from 0 to 7 T was used as a ferromagnetic-free purity test. Direct current (dc) variable temperature magnetic susceptibility measurements were collected in the temperature range 5-300 K under applied fields of 0.5 and 1 T. Magnetization data were acquired at 2 -10 K at fields 1, 4, and 7 T. Magnetic susceptibility data was corrected for diamagnetism of the sample, estimated using Pascal's constants, in addition to contributions from the sample holder and eicosane. The magnetic susceptibility data was collected multiple times until at least two different batches reproduced the data. Reduced magnetization data was modeled in PHI [1] according to the spin Hamiltonian: Ĥ = DSz2 + E(Sx2 -Sy2) + gµBS·B.

X-ray structure determinations.
Single crystals suitable for X-ray structure analysis were coated with deoxygenated Paratone N-oil and mounted in MiTeGen Kapton loops (polyimide). Data for compounds 2 -5 was collected at 100 K using synchrotron radiation at the Argonne National Laboratory Advance Photon Source, ChemMatCARS. A full dataset suitable for structure determination was collected at 100 K using 30 keV radiation. The single crystals of 2 -5 did not show decay during data collection. Data was collected using a Bruker three-circle platform goniometer equipped with a Bruker APEX II CCD and an Oxford Cryosystems cooling apparatus. The collection method involved 0.5° scans in ϕ at -5° in 2θ. Data integration down to at least 0.84 Å resolution was carried out using SAINT V8.34 C (Bruker diffractometer, 2014) with reflection spot size optimization. Absorption corrections were applied using SADABS. [2] The structure was solved by the Intrinsic Phasing methods and refined by least-squares methods again F 2 using SHELXL [3] with the OLEX 2 [4] interface. Space group assignments were determined by examination of systematic absences, E-statistics, and successful refinement of the structures. The program PLATON [5] was employed to confirm the absence of higher symmetry. Non-hydrogen atoms were refined with anisotropic displacement parameters, and hydrogen atoms were added in idealized positions and refined using a riding model. Crystallographic data for 4 and 5 are given in Table   S1.
The energy of the beamline at Argonne National Laboratory Advance Photon Source, ChemMatCARS, was referenced by performing a fluorescence scan of the Fe K-edge of a single crystal of sublimed ferrocene in steps of 1 eV using a Vortex-EX 3070 silicon drift diode detector.

S6
The ferrocene K-edge obtained this way was then compared to an XAS K-edge of sublimed ferrocene obtained at the Stanford Synchrotron Radiation Lightsource and the energy of the beamline was adjusted so that the energies of the pre-edge and rising edge as determined by both techniques were well aligned. All Fe XAS K-edge data collected at the Stanford Synchrotron Radiation Lightsource was referenced to iron foil.
Data collection strategy.
The following general procedure was followed for all samples: first, the crystal was mounted when the beam energy was at 30 eV. At this energy the crystal quality was judged by checking for the presence of twin domains. The crystal of each molecule which was both most single and gave the highest resolution at lowest exposure times was then selected to be used for the collection. A full structure diffraction dataset was then collected at this energy (30 keV).
Second, the beam energy was lowered to the Cr K-edge and a fluorescence scan (Figure 2c top and S4 top) was collected from 5939 to 6939 eV in steps of 1 eV by using a Vortex-EX 3070 silicon drift diode detector. This data was used to determine the energies at which to collect the partial diffraction data. Third, partial diffraction data was collected as described below at increasing energy.
For 2, 4, and 5, data was collected at 21 energies between 5967 and 6027 eV. For 3, data was collected at 18 energies between 5967 and 6027 eV. For 2-5, a total of 480 frames were collected at each energy using 4 pairs of φ scans with 2θ angles of -10°, -30°, -60°, and -90° for each pair and ω angles of 180° and 220° for the two scans within each pair. The step size in ϕ was 0.5° for all scans.
The 30 keV data was modeled to provide structural data on the compound of interest. This model was used as the reference for each partial diffraction dataset as is described below. For the data integration of the highest energy anomalous diffraction data set, the unit cell parameters of the reference (.p4p file) were imported into APEX3 to provide a starting point.  Tables S3-S8   Table S1. Crystallographic data for ( tbs L)Cr3(CNBn) (4) and ( tbs L)Cr3(µ 3 -NPh)(NPh) (5).
( tbs L)Cr3(CNBn) (4) ( tbs L)Cr3(µ 3 -NPh)(NPh) (5)    Figure S8. Variable-temperature magnetic susceptibility of 2. Data collected on heating from 5 to 292 K. A fit of the ground spin state is shown in solid black, while a linear extrapolation of this fit to higher temperatures in shown as a dashed black line. The fit corresponds to a single spin system with S=1, g = 1.99, and D = 5.64. The peak near 50 K is caused by a minor impurity of oxygen in the sample chamber. These points were excluded from the fit, as were the two data points at higher temperature which have anomalously low values due to instrument error. Figure S9. Variable-temperature magnetic susceptibility of 3. Data collected on heating from 5 to 300 K. Note that the expected spin-only value of the magnetic susceptibility for a triplet ground state is 1 while the expected value for a singlet ground state is 0. The missing points have highly negative values (-5 to -7) due to instrument error.

S21
S23 Figure S10. Variable-temperature magnetic susceptibility of 5. Data collected on heating from 5 to 292 K. A fit of the ground spin state is shown in solid black, while a linear extrapolation of this fit to higher temperatures in shown as a dashed black line. The fit corresponds to a single spin system with S=1, g = 2.01, and D = -6.10. Figure S11. UV-Vis/Near-IR spectra of 2 (green), 3 (blue), 4 (red), and 5 (purple) at 298 K in THF. The region between 5800 and 6000 cm -1 contains a strong signal from THF solvent background that saturated the detector and has therefore been omitted. Figure S12. Overalay of edge-scan derived predicted f'spectra for mononuclear references with the average f' of ( tbs L)Cr3(CNBn) (gray). Intensity of the gray trace was set to match the minimum of the trace for the Cr(IV), Cr(III), and Cr(II) references, respectively, in (a)-(c). This is included to demonstrate that due to the arbitrary intensity of the predicted f' spectra it is not possible to productively compare the location of the falling or rising edges between these predicted f' spectra and the experimentally collected f' data.
S26 Figure S13. Bond metrics for the imido fragment of 2. Figure S14. Bond metrics for the imido fragments of 5.