An Air‐Stable, Neutral Phenothiazinyl Radical with Substantial Radical Stabilization Energy

Abstract The vital effect of radical states on the pharmacological activity of phenothiazine‐based drugs has long been speculated. Whereas cationic radicals of N‐substituted phenothiazines show high stability, the respective neutral radicals of N‐unsubstituted phenothiazines have never been isolated. Herein, the 1,9‐diamino‐3,7‐di‐tert‐butyl‐N 1,N 9‐bis(2,6‐diisopropylphenyl)‐10H‐phenothiazin‐10‐yl radical (SQH2 .) is described as the first air‐stable, neutral phenothiazinyl free radical. The crystalline dark‐blue species is characterized by means of EPR and UV/Vis/near‐IR spectroscopy, as well as cyclic voltammetry, spectro‐electrochemical analysis, single‐crystal XRD, and computational studies. The SQH2 . radical stands out from other aminyl radicals by an impressive radical stabilization energy and its parent amine has one of the weakest N−H bond dissociation energies ever determined. In addition to serving as open‐shell reference in medicinal chemistry, its tridentate binding pocket or hydrogen‐bond‐donor ability might enable manifold uses as a redox‐active ligand or proton‐coupled electron‐transfer reagent.


Experimental Details
All reagents and solvents were purchased from commercial sources and were used as received unless otherwise noted. All used solvents were degassed prior to use either by three freeze-pumpthaw cycles or by ultrasonication with short vacuum application and were stored over activated molecular sieve (3 or 4 Å) under dry argon. All reactions were carried out in flame-dried standard laboratory glassware under a dry argon atmosphere using Schlenk line techniques and were permanently magnetically stirred. All syringes, magnetic stirring bars, needles, and transfer cannulas were dried and/or flushed with argon prior to use. Compounds sensitive to ambient conditions were stored and handled in a glove box (MBraun LABmaster dp, MB-20-G) filled with dry nitrogen gas. Removal of solvents or other volatiles in vacuo was performed using a Heidolph VV2000 rotary evaporator or a Schlenk line. Reported yields refer to isolated yields of analytically pure material and are the result of one specific reaction. Compounds known to literature were synthesized following published procedures, which are cited. Analytical data of literature-known compounds were compared to data of the respective reference and were found to be consistent in all cases.
Novel compounds were characterized to the reported structure to best of our knowledge.

Reactivity Studies
The reactivity of • with potential hydrogen atom donors was studied. This allowed us to narrow down the N-H bond strength with respect to homolytic bond cleavage. All reactions were carried out following the described procedure except otherwise noted.
In a J. Young NMR tube, • (3 mg, 4.5 μmol, 1.0 eq) was dissolved in dry and degassed DCM-d2 and the respective substrate was added (22.5 μmol, 5.0 eq). The reaction was followed for several days at room temperature using H NMR and EPR spectroscopy. S-16 S-17

Evans NMR Study
To follow the synthesis of • and to determine its effectiv magnetic moment, an in situ Evans NMR experiment [7] was carried out. For that, the precursor of • , the dihydro dibromide salt

EPR Spectroscopy
Electron paramagnetic resonance spectroscopy of • and • dissolved in dichloromethane was carried out with a Magenttech Miniscope MS 400 spectrometer at 298 K in the X band frequency range. Simulations of EPR spectra was done with Winsim v.1.0.

X-Ray diffraction analyses
Suitable crystals of the compounds were taken directly out of the mother liquor, immersed in perfluorinated polyether oil, and fixed on top of a cryo loop. Measurements were made on a Nonius-Kappa charge-coupled device diffractometer with a low-temperature unit using graphite-monochromated Mo-Kα radiation or on a Bruker APEX-III CCD diffractometer diffractometer with a lowtemperature unit using Mo-Kα radiation, chromated by mirror optics. The temperature was set at 120 K (Nonius-Kappa diffractometer) or 100 K (Bruker APEX-III CCD diffractometer). The data collected were processed using standard Nonius software [8] or Bruker APEX3 software. [9] Structures were solved by direct methods using the SHELXS or by intrinsic phasing using the SHELXT S-24 program and refined with the SHELXL. [10] Graphical handling of the structural data during solution and refinement was performed with Olex2 v1.3. [11] Atomic coordinates and anisotropic thermal parameters of non-hydrogen atoms were refined by full-matrix least-squares calculations [10c, 12] and displacement ellipsoides are displayed with 50 % probability. Hydrogen atoms were included using a riding model.

Computational Details
All quantum chemical calculations were performed with Orca 4.0.1 [13] using the computational resources of the bwUniCluster at the Karlsruhe Institute of Technology (KIT) within the Baden-Württemberg High Performance Computing program (bwHPC).
For all density functional theory calculations the chain of spheres exchange approximation (COSX) [14] combined with the Split-RI-J algorithm [15] (RIJCOSX) was used. Also, the D3 version of Grimme's empirical dispersion correction [16] along with Becke-Johnson damping [17] was always included. All molecular structures were optimized using the meta-GGA TPSS functional [18] with the def2-TZVPP basis set [19] as implemented in

Radical Stabilization Energy
Radical stabilization energies (RSE) were computed following the scheme from HIOE et al. [22] N H H N H The standard reaction enthalpy of the reference reaction shown above represents the difference in N-H bond dissociation energy and therefore is interpreted as the radical stabilization energy.