Ultrafast photophysics of para-substituted 2,5-bis(arylethynyl) rhodacyclopentadienes: thermally activated intersystem crossing

2,5-Bis(phenylethynyl) rhodacyclopentadienes (RCPDs), as a type of Rh(iii) complex, exhibit unusually intense fluorescence and slow intersystem crossing (ISC) due to weak metal–ligand interactions. However, details on their ultrafast photophysics and ISC dynamics are limited. In this work, electronic relaxation upon photoexcitation of two substituted RCPDs with two –CO2Me (A-RC-A) or –NMe2/–CO2Me (D-RC-A) end groups are comprehensively investigated using femtosecond transient absorption spectroscopy and theoretical analysis. Upon ultraviolet and visible excitation, dephasing of vibrational coherence, charge transfer, conformation relaxation, and ISC are observed experimentally. By calculating the spin–orbit coupling, reorganization energy, and adiabatic energy gap of plausible ISC channels, semi-classical Marcus theory revealed the dominance of thermally activated ISC (S1 → T2) for both D-RC-A and A-RC-A, while S1 → T1 channels are largely blocked due to high ISC barriers. With weak spin–orbit coupling, such differences in plausible ISC channels are predominately tuned by energetic parameters. Singlet oxygen sensitization studies of A-RC-A provide additional insight into the excited-state behavior of this complex.


General information
The complexes D-RC-A and A-RC-A were synthesized according to a previous report. [1]For all measurements, samples were dissolved in spectroscopic-grade solvents purchased from Sigma-Aldrich and used as received.The static UV/visible absorption spectra were performed on a JACSO V-670 spectrometer or an Agilent Cary 60 spectrophotometer using standard 1 cm pathlength quartz cells.Excitation and emission spectra were recorded at right angles to the excitation source on an Edinburgh Instruments FLS1000 spectrometer, equipped with a 450 W Xenon arc lamp, double monochromators for the excitation and emission pathways, and a red-sensitive photomultiplier (PMT-980) as detector.The excitation and emission spectra were corrected using the standard corrections supplied by the manufacturer for the spectral power of the excitation source and the sensitivity of the detector.Photoluminescence quantum yields measurements were performed by using a Quantaurus C11347 integrated sphere (Hamamatsu, Japan) exciting the sample at λex between 292 nm and 515 nm.The luminescence lifetime was measured using a FluoTime 300 spectrometer from PicoQuant equipped with a double-grating excitation monochromator, diode lasers (operating at 317 nm, 440 nm, 505 nm, pulse width < 80 ps) operated by a computer-controlled laser driver PDL-828 "Sepia II" (repetition rate up to 80 MHz, burst mode for slow and weak decays), two double-grating emission monochromators along with a PicoHarp 300 detector for TCSPC measurements (minimum base resolution 4 ps).The instrument response function calibration (IRF) was recorded using a diluted Ludox® dispersion.Lifetime analysis was performed using the commercial EasyTau 2 software (PicoQuant).The quality of the fit was assessed by minimizing the reduced chi squared function (χ 2 ) and visual inspection of the weighted residuals and their autocorrelation.

Femtosecond transient absorption measurements
The details of the employed fs-TA setup have been described elsewhere. [2,3]Briefly, a commercial Ti:sapphire amplifier (Solstice, Spectra-Physics) at 1 kHz repetition rate was employed as the main laser source.The 120 fs pulses at 800 nm were converted to λex = 513 nm and 295 nm as pump pulses for λex-dependent fs-TA experiments by using a non-collinear optical parametric amplifier (TOPAS White, Light Conversion) and subsequent up-conversion with the fundamental beam.A duration of ~25 fs was extracted for visible pulses by performing frequency-resolved optical gating (FROG) with type-I second-harmonic generation in BBO.For UV pulses (~105 fs), cross-correlation FROG with well-characterized fundamental pulses (800 nm, 100 fs) was employed by type-I difference-frequency generation (DFG) in BBO.The broadband probe pulses were generated by focusing 800 nm pulses into a linearly moving CaF2 window (2 mm thick).The linearly polarized pump and probe pulses were spatially overlapped in a 0.2 mm thick quartz cuvette with 54.7° mutual polarizations.
The delay time was varied up to 3.8 ns by using a commercial translation stage (M-IMS600, Newport).The probe pulses were eventually dispersed in a spectrometer (Acton SP2500i, Princeton Instruments) and detected shot-to-shot by a CCD camera (Pixis 2K, Princeton Instruments).The transient data measured were evaluated via target analysis [4] with the software package Glotaran based on the R-package TIMP. [5]

Theoretical calculations
All electronic structure calculations of D-RC-A and A-RC-A on ground and excited states were performed with the Gaussian 16 package. [6]For the S0, S1 and T2 states of D-RC-A and A-RC-A, a time-dependent density functional theory (TD-DFT) approach was employed for excitation energy calculations and geometric optimizations, while unrestricted density functional theory (UDFT) calculations were performed for optimizing the T1 states. [7,8]For light elements, PBE0/6-31g* level was employed while LANL2DZ effective core potential basis set was used for Rh(III). [9,10]The natural transition orbital (NTO) analysis of low-lying excited states was performed using the Multiwfn program, [11] and resulting NTO distributions were visualized by the VMD program. [12]MOMAP software [13][14][15] was employed to estimate the reorganization energy (Γ) of corresponding transitions with a harmonic oscillator approximation, [16,17] . ( [20] The Δqk can be further expressed as a linear combination of internal coordinates, i.e., , in which ΔDj represents the displacement with respect to the equilibrium position along internal coordinate j.The SOC matrix elements S1|ĤSO|T1 and S1|ĤSO|T2 were calculated by using a linear response approach implemented in the PySOC program, [21] in which the SOC Hamiltonian can be approximately described as where L and S represent magnetic moments resulting from orbital and spin angular momentum with the SOC constant (ri).

S4. Estimation of the ISC kinetics of A-RC-A under VIS excitation
Because the S1 state experiences a parallel reaction contributed to by three first-order reactions, i.e., radiative transition, internal conversion, and ISC with rate constants kr, kIC, and kISC, respectively, the decay of the S1 state is also a first-order reaction with rate constant of kS1 = kr + kIC + kISC.Meanwhile, the quantum yield of ISC can be calculated as ΦISC = kISC / (kr + kIC + kISC) = kISC / kS1, which gives kS1 = kISC / ΦISC.Similarly, we have kS1 = kr / Φf and kS1 = kIC / ΦIC.
Upon UV excitation, the ISC rate of A-RC-A can be calculated as kISC UV = 1/τISC UV = 2.58  10 8 s -1 , in which the time constant of ISC (τISC UV = 3.88 ns) was determined by fs-TA.
For estimating the corresponding triplet-state quantum yield (ΦISC UV ), we assumed ΦISC UV is approximately equal to the maximum value of singlet oxygen quantum yield (Φ Δ ) sensitized by the triplet state of A-RC-A, which was determined as 0.34.Considering that ~11% of the singlet-state oxygen will be quenched by A-RC-A in toluene-d8, the value of ΦISC UV can be deduced proportionally as ~0.37.So that the decay rate of S1 state can be calculated as kS1 UV = kISC UV / ΦISC UV = 6.96  10 8 s -1 , the S1 state lifetime is τS1 UV = 1/ kS1 UV = 1.44 ns, which is highly consistent with the measured fluorescence lifetime (~1.6 ns).
For estimating the ISC rate of A-RC-A upon visible excitation, we assumed that the slowing down of ISC upon visible excitation can be attributed to different S1-state decay rate, then kISC VIS = kISC UV + kS1 VIS -kS1 UV = 1.74  10 8 s -1 , while the ISC time constant is τISC VIS = 1/kISC VIS = 5.75 ns.

S5. Studies of 1 O2 sensitization by A-RC-A
Figure S14.Photobleaching of A-RC-A in air-saturated toluene-d8 solution before (black) and after (red) 10 min irradiation at 417 nm (fs laser operated at 0.5 kHz repetition, average power of 1.8 mW/cm 2 ).The solution was not stirred therefore we cannot quantify this change in absorbance.However, the same experiment in toluene-h8 did not show evidence of bleaching.
Therefore, given the appreciably longer lifetime of singlet oxygen in toluene-d8, and hence the increased probability of reaction with a solute, we infer that A-RC-A bleaching under these conditions is due to singlet oxygen.obtained from the single exponential fits are consistent with the expectation for toluene-h8. [22]

S6. NMR Spectra and Purification of A-RC-A
The compound was purified by column chromatography (Al2O3, basic) with n-hexane/THF 30:1 to 0:1 (6 or 7 times).The BHT impurity was reduced by column chromatography (Al2O3, basic) flushing with 10 volumes of toluene, and recovering the product in pure THF.
During the course of our current study, we noticed that the NMR spectra and data for A-RC-A presented in our previous paper were likely that for a different derivative which was not discussed in that paper. [1]In particular, the position of the signals for the Me group of the terminal CO2Me fragment in the 1 H and 13 C NMR spectra were upfield from where they would be expected (cf: the directly analogous compound with a (CH2)4-containing backbone).Thus, we re-recorded the NMR spectra and the new, correct, spectra and data are provided below.
We note that the original elemental analysis, mass spectroscopic data, and single-crystal Xray diffraction data are all correct as are all other NMR spectra and data in the previous paper.
In addition, the newly recorded absorption and emission spectra and fluorescence lifetime data are in agreement with the data presented in the previous paper. [

Figure S4 .
Figure S4.Time traces (open circles) and fitting curves (solid lines) by target analysis of the fs-TA at selected probe wavelengths for D-RC-A upon UV (a) and visible (b) excitation, as well as A-RC-A upon UV (c) and visible (d) excitation.

Figure S5 .S12Figure S6 .
Figure S5.Target-analysis-extracted concentration evolution of transient species for D-RC-A upon UV (a) and visible (b) excitation, as well as A-RC-A upon UV (c) and visible (d) excitation.

Figure S8 .
Figure S8.TD-DFT-calculated reorganization energy of each vibrational mode for the S1→S0 transition of D-RC-A (a) and A-RC-A (b).The values of total reorganization energy are provided for each complex and transition.

Figure S9 .Figure S10 .
Figure S9.Chirp-corrected spectro-temporal maps of fs-TA signal in -0.5 -3.0 ps range for D-RC-A (a) and A-RC-A (b) upon visible optical excitation at λex = 513 nm; the corresponding TA spectra at selected delay times are illustrated in (c) and (d), respectively.

Figure S11 .
Figure S11.Emission decay of A-RC-A in THF collected at 317 nm excitation.A lifetime of 1.6 ns was determined (χ 2 = 1.02).

Figure S12 .
Figure S12.Emission decay of A-RC-A in THF collected at 440 nm excitation.A lifetime of 1.6 ns was determined (χ 2 = 1.01).

Figure S13 .
Figure S13.Emission decay of A-RC-A in THF collected at 505 nm excitation.A lifetime of 1.6 ns was determined (χ 2 = 1.02).

Figure S15 .
Figure S15.Time-resolved O2(a 1 Δg) phosphorescence traces recorded upon pulsed laser irradiation of A-RC-A at 417 nm in toluene-h8 for different laser powers (fs laser operated at 1 kHz repetition).Single exponential fits to the data are shown as solid lines.The traces recorded at 4.5 and 4.2 mW overlap appreciably and are almost indistinguishable from each other.In all cases, an intense "spike" coincident with the laser pulse, and likely deriving from A-RC-A fluorescence combined with scattered laser light, was detected at time = 0. We eliminated the data showing this spike for presentation in the Figure.The O2(a 1 Δg) lifetimes

Figure S16 .
Figure S16.Integrated intensity of the O2(a 1 Δg) phosphorescence signal, normalized by the sensitizer absorbance and the O2(a 1 Δg) lifetime, plotted as a function of laser power for A-RC-A and for the reference standards, phenalenone (PN) and tetraphenylporphyrin (TPP).In such plots, the slopes of the linear fits are proportional to the O2(a 1 Δg) quantum yield.

Figure S17 .
Figure S17.Representative time-resolved O2(a 1 Δg) phosphorescence traces recorded upon irradiation of PN at 417 nm in toluene-d8.Data were recorded as a function of the O2 concentration, controlled by the percent of oxygen in a mixture of O2 and N2 gas bubbled through the solvent.(a) The data from 100 µs to 2000 µs were fitted by a single exponential decay function to obtain the lifetime of O2(a 1 Δg) (i.e., 1/k Δ ).(b) Using k Δ as a fixed parameter, eq 5 was used as a fitting function to obtain values of kT for a time domain where O2(a 1 Δg) was formed in the photosensitized reaction.Fits are shown as solid lines superimposed on each trace.

Table S1 .
Reported photophysical properties of 2,5-substituted RCPDs with two electron acceptors (A-A), two electron donors (D-D), and electron donor/acceptor (D-A) in toluene under oxygen-free conditions.

Table S2 .
Calculated vertical and adiabatic transitions of the lowest-lying excited states (S1, T1 and T2) of D-RC-A and A-RC-A, H = HOMO, L = LUMO, CT = charge transfer, LE = local excited.

Table S3 .
TD-DFT-optimized angles and dihedral angles (numbering see FiguresS6 and S7) of D-RC-A and A-RC-A at the S0, S1, T1, and T2 states.