Luminescence and Light‐Driven Energy and Electron Transfer from an Exceptionally Long‐Lived Excited State of a Non‐Innocent Chromium(III) Complex

Abstract Photoactive metal complexes employing Earth‐abundant metal ions are a key to sustainable photophysical and photochemical applications. We exploit the effects of an inversion center and ligand non‐innocence to tune the luminescence and photochemistry of the excited state of the [CrN6] chromophore [Cr(tpe)2]3+ with close to octahedral symmetry (tpe=1,1,1‐tris(pyrid‐2‐yl)ethane). [Cr(tpe)2]3+ exhibits the longest luminescence lifetime (τ=4500 μs) reported up to date for a molecular polypyridyl chromium(III) complex together with a very high luminescence quantum yield of Φ=8.2 % at room temperature in fluid solution. Furthermore, the tpe ligands in [Cr(tpe)2]3+ are redox non‐innocent, leading to reversible reductive chemistry. The excited state redox potential and lifetime of [Cr(tpe)2]3+ surpass those of the classical photosensitizer [Ru(bpy)3]2+ (bpy=2,2′‐bipyridine) enabling energy transfer (to oxygen) and photoredox processes (with azulene and tri(n‐butyl)amine).


S1
Supporting Information General Procedures. Diethyl ether was distilled over sodium, THF over potassium and acetonitrile over calcium hydride. The ligand tpe was prepared similar to a reported procedure. S1 NMR spectroscopic and mass spectrometric data match the literature values. A glovebox (UniLab/MBraun, Ar 4.8, O2 < 100 ppm, H2O < 1 ppm) was used for storage and weighing of sensitive compounds. Reagents were received from usual suppliers (ABCR, Acros Organics, Alfa Aesar, Fischer Scientific, Fluka and Sigma Aldrich). NMR spectra of tpe were recorded on a Bruker Avance DRX 400 spectrometer at 400.31 MHz ( 1 H). All resonances are reported in ppm versus the solvent signal as internal standard [CDCl3 ( 1 H:  = 7.26)]. S2 IR spectra were recorded with a Bruker Alpha FTIR spectrometer with ATR unit containing a diamond crystal. ESI + mass spectra were recorded on a Micromass Q-TOF-Ultima spectrometer. DC magnetic studies were performed with a Quantum Design MPMS-XL-7 SQUID magnetometer on powdered microcrystalline samples. Experimental susceptibility data were corrected for the underlying diamagnetism using Pascal's constants. The temperature dependent magnetic contribution of the holder and of the embedding eicosane matrix was experimentally determined and subtracted from the measured susceptibility data. Variable temperature susceptibility data were collected in a temperature range of 6 -300 K under an applied field of 0.1 Tesla. Electrochemical experiments were carried out on a BioLogic SP-50 voltammetric analyzer using platinum wires as counter and working electrodes and a 0.01 M Ag/Ag[NO3] electrode as reference electrode. Cyclic voltammetry and square wave measurements were carried out at scan rates of 50-200 mV s -1 using 0.1 M [N n Bu4][BF4] in CH3CN as supporting electrolyte. Potentials are referenced against the ferrocene/ferrocenium couple. Spectroelectrochemical experiments were performed using a Specac omni-cell liquid transmission cell with CaF2 windows equipped with a Pt gauze working electrode, a Pt gauze counter electrode and a Ag wire as pseudo reference electrode, melt-sealed in a polyethylene spacer (approximate path length 0.5 mm) in 10 -5 M solutions in CH3CN, containing 0.1 M [N n Bu4] [BF4]. UV/Vis/NIR spectra were recorded on a Varian Cary 5000 spectrometer using 1.0 cm cells. Luminescence emission spectra and decays in solution were reported with a calibrated spectrofluorometer FSP 920 from Edinburgh Instruments. For the measurement of the emission spectra, a continuous xenon lamp was applied as excitation light source, while the time-resolved luminescence measurements were completed with a µs xenon flashlamp and detection in a multi-channel scaling mode. All measurements were performed at magic angle condition (polarization 0° in the excitation and 54.7° in the emission channel). The luminescence decays in solution were analyzed by fitting the obtained decay curves mono-exponentially with the program FAST (Fluorescence Analysis Software Technology, Edinburgh Instruments Ltd.). The luminescence quantum yields in solution were determined using an Ulbricht integrating sphere (Quantaurus-QY C11347-11, Hamamatsu). S3-S5 Relative uncertainty is estimated to be ±5 %. NIR absorption spectra of CH3CN and CD3CN for overtone determination were measured in absorption mode using a JASCO V-770 spectrophotometer (equipped with a long cuvette holder LSE-701). The spectra were recorded in rectangular semimicro cuvettes (Starna, type 28/B/SX/50, path length 5.0 cm). CH3CN was spectrophotometric grade and CD3CN was NMR grade (99.8% D). The spectra were corrected for baseline drift with OriginPro 9.0. The component peaks of the spectra were deconvoluted by fitting of the spectra with a series of Gaussian functions (Levenberg-Marquardt on  2 ). All timeresolved FTIR experiments were performed with an FTIR spectrometer Bruker Vertex 80v, operated in the step-scan mode. KBr pellets of [Cr(tpe)2][BF4]3 (ca. 0.75 mg) were prepared by mixing with dry KBr (ca. 200 mg, stored at 80 °C) and grinding to a homogeneous mixture. The strongest peak in the ground state spectrum showed an absorption of about 0.6 OD with the mentioned concentration. Measurements with cryogenically cooled KBr pellets (20 K and 290 K at the sample) were performed with a closed cycle helium cryostat (ARS Model DE-202A). The cryo cooler was equipped with a homebuilt pellet holder and CaF2 windows. A liquid-nitrogen-cooled mercury cadmium telluride (MCT) detector (Kolmar Tech., Model KV100-1-B-7/190) with a rise time of 25 ns, connected to a fast preamplifier and a 14-bit transient recorder board (Spectrum Germany, M3I4142, 400 MS s −1 ), was used for signal detection and processing. The laser setup includes a Q-switched Nd:YAG laser (Innolas SpitLight Evo I) generating pulses with a band-width of 6 -9 ns at a repetition rate of 100 Hz. The third harmonic (355 nm) of the Nd:YAG laser was used for sample excitation. The UV pump beam was attenuated to about 2.0 mJ per shot at a diameter of 9 mm. The beam was directed onto the sample and adjusted to have a maximal overlap with the IR beam of the spectrometer. The sample chamber was equipped with anti-reflection-coated germanium filters to prevent the entrance of laser radiation into the detector and interferometer compartments. The time delay between the start of the experiment and the UV laser pulse was controlled with a Stanford Research Systems DG535 delay generator. A total of more than 2000 coadditions were recorded at each interferogram point. The spectral region was limited by undersampling to 988 -1975 cm −1 with a spectral resolution of 4 cm −1 resulting in 555 interferogram points. An IR broad band filter (850 -1750 cm −1 ) and the CaF2 windows (no IR transmission < 1000 cm −1 ) of the cryostat prevented problems when performing a Fourier transformation (i.e. no IR intensity outside the measured region should be observed). FTIR ground state spectra were recorded systematically to check for sample degradation. A more detailed description of the step-scan setup is given here. S6-S8 Temperature dependent emission spectra of KBr disks were recorded with a FluoroMax-2 (Horiba Scientific) device using the described cryostat. Time-correlated single photon counting measurements were performed with a DeltaFlex (Horiba Scientific) instrument at a repetition rate of 10 kHz with a time resolution of 13 or 27 ns, depending on the experiment. Temperature dependent emission spectra of crystals of [Cr(tpe)2][BF4]3 and [Cr(tpe)2][PF6]3 were recorded with a Renishaw Invia Raman microscope equipped with a Peltier-cooled CCD camera. The excitation source was a 488 nm Argon ion laser. Variable-temperature spectra were obtained by coupling a Linkam cryostat to the microscope with liquid nitrogen used as the coolant. Elemental analyses were conducted by the microanalytical laboratory of the chemical institutes of the University of Mainz.
Caution! Although we have not experienced any problems in handling the perchlorate solutions, all materials should be handled with extreme care.
Crystal structure determinations. Diffusion of diethyl ether into concentrated solutions of [Cr(tpe)2][BF4]3 or [Cr(tpe)2][PF6]3 in CH3CN yielded diffraction quality crystals. Intensity data were collected with a STOE IPDS-2T diffractometer and an Oxford cooling system and corrected for absorption and other effects using Mo K radiation ( = 0.71073 Å). The diffraction frames were integrated using the SAINT package, and most were corrected for absorption with MULABS. S9,S10,S11 The structures were solved by direct methods and refined by the full-matrix method based on F 2 using the SHELXTL software package. S12,S13 All non-hydrogen atoms were refined anisotropically, while the positions of all hydrogen atoms were generated with appropriate geometric constraints and allowed to ride on their respective parent carbon atoms with fixed isotropic thermal parameters. CCDC Density functional theoretical calculations on the chromium complex cations [Cr(tpe)2] n+ were carried out using the ORCA program package (version 4.0.1). S14 Tight convergence criteria were chosen for all calculations (keywords tightscf and tightopt). All calculations make use of the resolution of identity (Split-RI-J) approach for the Coulomb term in combination with the chain-of-spheres approximation for the exchange term (COSX). S15,S16 Geometry optimization was performed using the B3LYP functional S17 in combination with Ahlrichs' split-valence triple- basis set def2-TZVPP for all atoms. S18,S19 The optimized geometries were confirmed to be local minima on the respective potential energy surface by subsequent numerical frequency analysis (Nimag = 0). TD-DFT calculations were performed at the same level of theory. Fifty vertical spin-allowed transitions were calculated. The zero order relativistic approximation was used to describe relativistic effects in all calculations (keyword ZORA). S20,S21 Grimme's empirical dispersion correction D3(BJ) was employed (keyword D3BJ). S22,S23 To account for solvent effects, a conductor-like screening model (keyword CPCM) modeling acetonitrile was used in all calculations. S24 Explicit counterions and/or solvent molecules were neglected.
Synthesis of 1,1,1-tris(pyrid-2-yl)ethane tpe: S1 2-Ethylpyridine (2.1 ml, 18.7 mmol) in dry THF (60 ml) were cooled to -78°C. n Butyl lithium (2.5 M in hexane; 7.5 ml; 18.75 mmol) was added dropwise. After stirring for 50 min and warming to -40°C, 2-fluoropyridine (3.6 g. 37.4 mmol) was added to the red solution with the temperature kept below -30°C. The resulting colorless reaction mixture was warmed to room temperature and then heated under reflux for 12 h. After cooling to room temperature, the solvents were removed under reduced pressure. The resulting solid was dissolved in THF (60 ml) and filtered. The solvent was removed under reduced pressure and the brown raw product was purified by column chromatography (alumina, hexanes/ethyl acetate 2:1, Rf = 0.30) giving tpe as off-white solid

Synthesis of [Cr(tpe)2][BF4]3:
Under oxygen-free conditions, tpe (500 mg, 1.91 mmol) was dissolved in a CH3CN/H2O mixture (50 ml, 1:1 v/v) and CrCl2 (118 mg, 0.96 mmol) was added as a solid. The resulting dark green solution was stirred for 2 h at room temperature, heated to reflux for 3 h and stirred for 5 d at room temperature. [NH4][BF4] (312 mg, 2.98 mmol) dissolved in deaerated water (3 ml) was added to the reaction mixture. After stirring for 2 h under inert conditions, the mixture was left to stand for 16 h under air giving a pale-red solution. The solvents were removed under reduced pressure and the resulting red solid suspended in CH3CN (5 ml  Both salts crystallize with three acetonitrile solvate molecules according to single crystal XRD analyses (see below). However, the actual amount of solvent in a given sample depends on the grinding and drying procedure after crystallization.   has been studied in the range 6 -300 K in an external field of 0.1 T. In this temperature range, T is essentially temperature independent and with 1.875 cm 3 K mol -1 very close to the spin-only value of a S = 3 /2 system ( 4 A2g ground state; g = 2.000; µ = 3.87 µB). The fitted g value matches those of [Cr (ddpd)2     S18 Figure S18.
Step-scan FT-IR spectrum (red) and ground state IR spectrum ( Figure S19. Decay curves and global fits obtained from step-scan FT-IR data at a) 290 K and b) 20 K including residuals. c) -f) Fits of the transition around 1465 cm -1 at 290 K and 20 K (positive band that belongs to the excited state) and transition around 1470 cm -1 at 290 K and 20 K (negative band that belongs to the electronic ground state) including residuals. Figure S20. Decay curves as well as a) tri-and b) tetraexponential fits obtained from TCSPC data at 290 K (obs = 722 nm, KBr). The offset results from the dark current.
a) b) Figure S21. Decay curves as well as a) tri-and b) tetraexponential fits obtained from TCSPC data at 290 K (obs = 733 nm, KBr). The offset results from the dark current.
a) b) Figure S22. Decay curves as well as a) tri-and b) tetraexponential fits obtained from TCSPC data at 290 K (obs = 745 nm, KBr). The offset results from the dark current. Figure S23. a) Emission spectra of [Cr(tpe)2][BF4]3 as neat film at 290 K with different excitation wavelengths, decay curves as well as b) tri-and c) tetraexponential fits obtained from TCSPC data at 290 K (obs = 717 nm), decay curves as well as d) tri-and e) tetraexponential fits obtained from TCSPC data at 290 K (obs = 740 nm). Figure S24. Decay curves and biexponential fits obtained from TCSPC data at 20 K (obs = 740 nm, KBr). Figure S25. Decay curves and biexponential fits obtained from TCSPC data at 20 K (obs = 759 nm, KBr).

Spectral overlap integral (SOI) calculations for [Cr(tpe)2] 3+ / CH3CN and CD3CN:
The C-H overtone bands of CH3CN in solution ( Figures S26 and S27) were measured in the relevant spectral region (11500 to 16000 cm -1  fourth C-H stretch overtone and third combination region) corresponding to the spectral region of the doublet emission of [Cr(tpe)2] 3+ . The obtained data are consistent with previous measurements in liquid CH3CN and show that the band intensities are largely dominated by C-H stretching overtones without significant participation of CN vibrations. S28 The measured bands were fitted by a series of Gaussian functions in order to extract a coherent expression of the band shape for the SOI calculation (vide infra).

S26
The C-D absorption bands in the same region as for the C-H overtones (11500 to 16000 cm -1 ) could not be reliably measured due to the extremely small molar extinction values associated with these overtones. In order to be able to calculate the corresponding SOI with the chromium emission, the spectral characteristics of the required higher overtones C-D (in particular the overtones  = 6 and 7) were extrapolated from the systematic changes seen in the lower overtones ( = 3 -5) that could be measured. For this purpose, each measured overtone band was fitted with a single Gaussian function of the form y(x) = A  exp(-0.5*((x-xc)/) 2 ) defined by amplitude A, center wavenumber xc, and Gaussian width .
As can be seen from Figures S28-S30, this is a rather crude approximation for the lower overtone bands (especially seen in Figure S28) but improves with increasing vibrational quantum number as the vibrations gain more local-mode character and coupling between oscillators is diminishing.   Due to the high degree of coupling in the overtones, the only Gaussian parameter that turned out to be reliable for the extrapolation was the amplitude A. As already successfully shown in previous work, S29 the semilogarithmic plot of log(A) vs. vibrational quantum number  could be fitted well by a linear relationship ( Figure S31). The other two Gaussian parameters (xc and ) necessary for the construction of the higher C-D overtone bands were estimated using the published data. The xc values were calculated using the known Morse parameters for liquid CD3CN (fundamental 0  = 2175 cm -1 and anharmonicity x = 30.5 cm -1 ), S30 whereas the Gaussian widths for the required overtones  were estimated to be equal to the corresponding C-D overtone band widths for aromatic C-D oscillators. S29,S31 The properties of the S28 measured and extrapolated Gaussians are summarized in Table S4 and were used for the SOI calculations (vide infra). [a] Based on Gaussian fits of measured data.   Based on the obtained SOIs, the ratio of the non-radiative deactivation rates of chromium excited states in [Cr(tpe)2] 3+ in the acetonitrile isotopologues is estimated as: S31   -21  -1  3  3  nr  3  3  -25  -1  3  nr  3  3 k (CH CN) SOI(CH CN) 1.24 10 M cm = = = 1.97 10 k (CD CN) SOI(CD CN) 6.28 10 M cm

  
This ratio would suggest a massive reduction in multiphonon relaxation rates and a concomitant increase in luminescence intensity and lifetimes by exchanging CH3CN versus CD3CN. Yet, this is not observed. For a discussion, see main text.