Stereoelectronic Effects in Cl2 Elimination from Binuclear Pt(III

Oct 31, 2016 - First, using a method developed during a study of the solid-state photochemistry of mer-3,(14) evolved Cl2 was exposed to a strip of Na...
1 downloads 12 Views 815KB Size
Article pubs.acs.org/IC

Stereoelectronic Effects in Cl2 Elimination from Binuclear Pt(III) Complexes David C. Powers,†,‡ Seung Jun Hwang,† Bryce L. Anderson,† Haifeng Yang,‡ Shao-Liang Zheng,† Yu-Sheng Chen,§ Timothy R. Cook,⊗ François P. Gabbaï,‡ and Daniel G. Nocera*,† †

Department of Chemistry and Chemical Biology, Harvard University, 12 Oxford Street, Cambridge, Massachusetts 02138, United States ‡ Department of Chemistry, Texas A&M University, College Station, Texas 77843, United States § ChemMatCARS, The University of Chicago, Argonne, Illinois 60439, United States ⊗ University at Buffalo, The State University of New York, Buffalo, New York 14260, United States S Supporting Information *

ABSTRACT: Halogen photoelimination is the critical energy-storing step of metal-catalyzed HX-splitting photocycles. Homo- and heterobimetallic Pt(III) complexes display among the highest quantum efficiencies for halogen elimination reactions. Herein, we examine in detail the mechanism and energetics of halogen elimination from a family of binuclear Pt(III) complexes featuring meridionally coordinated Pt(III) trichlorides. Transient absorption spectroscopy, steady-state photocrystallography, and far-infrared vibrational spectroscopy suggest a halogen elimination mechanism that proceeds via two sequential halogen-atom-extrusion steps. Solution-phase calorimetry experiments of the meridional complexes have defined the thermodynamics of halogen elimination, which show a decrease in the photoelimination quantum efficiency with an increase in the thermochemically defined Pt−X bond strength. Conversely, when compared to an isomeric facial Pt(III) trichloride, a much more efficient photoelimination is observed for the fac isomer than would be predicted based on thermochemistry. This difference in the fac vs mer isomer photochemistry highlights the importance of stereochemistry on halogen elimination efficiency and points to a mechanism-based strategy for achieving halogen elimination reactions that are both efficient and energy storing.



INTRODUCTION

Solid-state halogen elimination reactions have been demonstrated from mono- and binuclear complexes of Au15 and Pt,13,14 as well as from a family of mononuclear Ni(III) complexes.16,17 The relevance of stoichiometric X2 elimination reactions to energy storage has been experimentally established with thermochemical data for halogen elimination from mononuclear Pt13 and Ni16,17 complexes. Nonetheless, the definition of the thermodynamics of halogen elimination from a diverse set of molecular scaffolds and the relation of such data to the halogen elimination mechanism is lacking. As a part of our efforts toward understanding the structural factors that contribute to efficient halogen photoelimination, we have interrogated the Cl2-elimination photochemistry of a family of bimetallic Pt(III) trichlorides (Figures 1 and 2) with various ancillary metals.13,14,18,19 Herein, we establish highyielding, solid-state Cl2 elimination from homobimetallic meridionally coordinated Pt(III) trichloride complex mer-1. Nanosecond time-resolved transient absorption spectroscopy, carried out in the solution phase, and steady-state single-crystal photocrystallography measurements, accomplished in the solid state, implicate a photoelimination mechanism that proceeds

Photocatalytic splitting of hydrohalic acids to generate hydrogen and halogen equivalents (2 HX → H2 + X2) stores substantial energy, and constitutes a chemical basis for solar-tofuels energy storage technologies.1−5 In closed HX-splitting catalysis cycles, proton reduction and halide oxidation must be accomplished for turnover.6,7 Frequently, HX-splitting chemistry is thwarted by exothermic back reaction of photoevolved X2 with the reduced photocatalyst, and thus a productive photoreaction can only be achieved in the presence of chemical traps that sequester the photogenerated halogen.8−11 The use of chemical traps to sequester the energetic halogen product obviates energy storage by rendering the resulting photocycles energy neutral or exothermic.12 Authentic halogen elimination photoreactions, in which halogen elimination proceeds in the solid state in the absence of chemical traps, have been advanced as the foundation of energy-storing HX-splitting photocycles.7,13−15 In addition to supporting energy storage, authentic halogen elimination reactions provide a mechanistic test to differentiate bona fide halogen elimination reactions from formal elimination reactions driven by direct participation of halogen traps with the molecular excited states of metal halide complexes. © XXXX American Chemical Society

Received: August 3, 2016

A

DOI: 10.1021/acs.inorgchem.6b01887 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 1. Pt2[III,III] complex mer-1 participates in trap-free, authentic Cl2 photoelimination. P−N−P ligand: bis(bis(trifluoroethoxy)phosphino)methylamine (tfepma).

Figure 3. Cl2 evolved during the photolysis of mer-1 has been quantified by both (a) ion chromatography and (b) colorimetrically as its aminyl radical.

Second, evolved Cl2 was treated with N,N-diethyl-1,4-phenylenediamine sulfate (DPD)16 and the resulting DPD•+Cl− was quantified colorimetrically to give a 83% yield for Cl2 from mer1 (Figure S2). The Cl2 yields from these methods are based on starting material conversion determined by 31P NMR. In both experiments, spatially separated chemical traps were employed during the photolysis of binuclear complex mer-1, confirming trap-free Cl2 elimination from mer-1. Time-Resolved Photochemistry. Laser flash photolysis of Pt2[III,III] complex mer-1 was carried out to probe the presence of non-steady-state intermediates in the photoreduction of mer-1. Complex mer-1 was irreversibly consumed during the collection of nanosecond-resolved transient absorption (TA) spectra (excitation from a 355 nm laser), and thus data was collected under irreversible flow conditions. The prompt TA spectrum (40 ns delay) obtained by laser flash photolysis of Pt2[III,III] complex mer-1 displays spectral growths centered at 330 and 460 nm and ground-state bleaches centered at 310 and 385 nm (Figure 4a). The spectral growths

Figure 2. Heterobimetallic mer-Pt(III) complexes that participate in halogen photoelimination reactions. P−N−P ligand: bis(bis(trifluoroethoxy)phosphino)methylamine (tfepma); C−C−P ligand: o-(Ph2P)C6H4.

via two sequential halogen-atom extrusion steps. Solution-phase calorimetry measurements have defined the thermodynamics of halogen elimination from the available family of meridional Pt(III) trichloride. Consistent with a unified M−X homolysis mechanism for Cl2 elimination from meridional PtCl3 centers, the efficiency of halogen elimination decreases with increasing M−X bond strength. Comparison of the efficiency of halogen elimination from this set of mer-Pt(III) complexes with the efficiency of Cl2 elimination from an isomeric fac-Pt(III) trichloride, from which efficient two-electron Cl2 reductive elimination via a σ-complex has been proposed,18 reveals the importance of stereochemistry in halogen photoelimination. These observations point to the utility of stereoelectronic effects to access halogen photoelimination reactions that are simultaneously efficient and highly endothermic.



Figure 4. (a) Normalized absorption spectrum of mer-1 (red) and TA spectrum (black) obtained by laser flash photolysis of a THF solution of mer-1 (355 nm pump). (b) Single-wavelength kinetic trace of the evolution of the spectral bleach at 380 nm.

RESULTS Characterization and Quantification of Gaseous Photoproducts. In the original report of the solid-state photochemistry of the Pt2[III,III] complex mer-1, photogenerated Pt2[I,III] complex 2 was detected and quantified by 31P NMR and photoevolved Cl2 was trapped cryogenically and subsequently characterized by in-line mass spectrometry.13 Although this experiment established the photoevolution of Cl2, it did not quantify the Cl2 evolved. Two experiments have been pursued to quantify evolved Cl2 in the photolysis of mer-1 (Figure 3). First, using a method developed during a study of the solid-state photochemistry of mer-3,14 evolved Cl2 was exposed to a strip of Na metal and the resulting NaCl was quantified by ion chromatography. Using this method, the Cl2 yield from mer-1 was determined to be 86% (Figure S1).

do not correspond to Pt2[I,III] complex 2, and their presence indicates the formation of a non-steady-state species in the time-resolved experiment. The spectral features observed are long-lived; substantial spectral evolution is not observed over 300 μs (Figure 4b). Identical long-lived spectra were obtained by laser flash photolysis of PhH solutions of mer-1 (Figure S3). TA spectra have also been obtained for the heterobimetallic Pt(III) complexes mer-3, mer-5, and mer-7. TA spectra and single-wavelength kinetic traces are collected in Figures S4−S6. In all cases, non-steady-state intermediates were observed. Similar to the photointermediate derived from laser flash B

DOI: 10.1021/acs.inorgchem.6b01887 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry photolysis of complex mer-1, the photointermediate derived from PtRh complex mer-5 is long-lived (Figure S5). In contrast, the non-steady-state intermediates derived from PtSb complex mer-3 and PtTe complex mer-7 are transient (Figures S4 and S6): the prompt TA spectra of mer-3 and mer-7 indicate the presence of photointermediates. The final TA spectra display only signals attributable to steady-state photoreaction products. Photocrystallography. The photoinduced structural changes involved in halogen elimination chemistry from Pt2[III,III] complex mer-1 were probed by steady-state photocrystallography experiments. Photocrystallographic analysis is based on generating a photodifference map in which diffraction data collected in the dark is compared with diffraction data collected during steady-state irradiation.20−25 We have previously applied photocrystallography experiments to identify halogen-bridged intermediates in photoelimination from Rh2 complexes26 and to elucidate the stereoelectronic effects in M−X bond activation at Ni(III) trihalide complexes.16 Photocrystallography data was collected with single crystals of mer-1 cooled to 15 K. Irradiation was supplied by a 365 nm LED source. Refinement of the obtained photodifference map (Figure 5) revealed the presence of a photogenerated structural

solution phase experiments (Figure S8). These experiments suggest that similar intermediates are accessible in both the solution-phase and the solid-state photolysis reactions. It should be noted that the TA spectra were recorded at 23 °C while the photocrystallographic experiments were carried out at 15 K. Attempts to interrogate the solid-state photochemistry of complexes mer-3 and mer-5 were unsuccessful due to rapid loss of sample crystallinity during single-crystal photolysis experiments. Far-Infrared Spectroscopy. To explore the selective extrusion of equatorial Pt−Cl ligands that is apparent from the photocrystallographic data, we evaluated the relative strengths of apical and equatorial ligands by vibrational spectroscopy. The far-infrared spectra (FIR) of the binuclear Pt complexes discussed herein are expected to contain bands corresponding to Pt−X bond stretches.28−30 While the FIR region can be obscured by a large number of bands, we have been able to establish which stretches correspond to M−X stretching modes by comparing isostructural Pt2 chlorides and bromides. The spectra of Pt2(tfepma)2Cl2 (9) and Pt2(tfepma)2Br2 (10; X-ray data: Figure S9 and Table S3) display very similar bands between 700 and 300 cm−1, which, while not assigned to specific vibrational modes, are clearly shared between the structurally similar complexes (Figure 6a). A unique band in the

Figure 5. Thermal ellipsoid plots of photocrystallography results with photogenerated structure (solid) superimposed on dark structure of mer-1 (faded). Pt(1)−Cl(2): 3.10(7) Å (photogenerated); 2.326(2) Å (dark).

component that represents 4.8(5)% of the irradiated crystalline sample. The metrical parameters of the photogenerated structure show substantial and selective elongation of one of the equatorial Pt−Cl bonds: Pt(1)−Cl(2): 2.326(2) Å (dark); 3.10(7) Å (photogenerated). Due to the low population of the photogenerated structure, the difference map was solved in the same monoclinic unit cell as the dark structure, which enforces equivalence of Pt(1)−Cl(2) and Pt(1a)−Cl(2a). Attempts to increase the population of the photogenerated structure by prolonged irradiation or by higher irradiation power led to sample decomposition. The photogenerated structure was observed in a data set collected after irradiation was halted, indicating that the photogenerated structure is not reversibly generated at 15 K. Warming the photolyzed crystal to 100 K led to sample decomposition, which prevented investigation of the thermal reversibility of the observed photochemistry. The photocrystallography experiment was reproduced with three unique crystalline samples of complex mer-1. Variable-temperature (VT) crystallography of single crystals of mer-1 revealed that the metrical parameters of mer-1 do not substantially vary with temperature, confirming that laserinduced heating of the sample is not responsible for the observed photogenerated structures.27 VT X-ray data is collected in Figure S7 and Tables S1 and S2. TA spectra obtained by laser flash photolysis of thin films of mer-1 display similar line shapes as spectra obtained during

Figure 6. FIR spectra of (a) 9 (black line) and 10 (red line) obtained with crushed crystalline samples at 23 °C; (b) 2 (black line) and 11 (red line) obtained with solid samples at 23 °C; and, (c) mer-1 (black line) obtained with a crushed crystalline sample collected at 23 °C.

spectrum of 9 at 264 cm−1 is absent in the spectrum of 10, and is replaced by a new band at 192 cm−1. This band is assigned to a Pt−X stretching mode. Both Pt2X2 complexes have idealized D2h symmetry and analysis of the two Pt−X vectors in this point group gives two vibrational modes, the fully symmetric Ag and asymmetric B3u, of which only the B3u mode is IR active. Assuming an identical force constant for the Pt−Cl and Pt−Br stretch, the expected ratio of νCl/νBr is ∼1.5, which is in fair agreement with the observed ratio of 1.375, considering that an identical force constant is a simplification in this system. The spectra of Pt2(tfepma)2Cl4 (2) and Pt2(tfepma)2Br4 (11; X-ray data: Figure S10 and Table S4) are similar to the spectra C

DOI: 10.1021/acs.inorgchem.6b01887 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry of Pt2X2 complexes between 700 and 300 cm−1 (Figure 6b). As observed for the Pt2[I,I] complexes, the unique bands of Pt2Cl4 complex 2, found at 267 and 240 cm−1, shift to 234 and 187 cm−1 for Pt2Br4 complex 11. The Pt2X4 molecules have idealized C2v symmetry, which predicts 3A1 and 1B1 vibrational modes involving the Pt−X vectors, all of which are IR active. Experimentally, only two bands are resolved, suggesting either spectral overlap of the 3A1 bands, or frequencies falling below the lower limit of the spectrometer (