REVIEW Multi-Photon Excitation in Photoredox Catalysis: Concepts, Applications, Methods Felix Glaser,+[a] Christoph Kerzig,+[a] and Oliver S. Wenger*[a] + These two authors contributed equally. 1 REVIEW [a] F. Glaser, Dr. C. Kerzig, Prof. Dr. O. S. Wenger Department of Chemistry University of Basel St. Johanns-Ring 19, 4056 Basel (Switzerland) E-mail: [email protected] Abstract: The energy of visible photons and the accessible redox usually formed in micro-molar concentrations.[3] The potentials of common photocatalysts set thermodynamic limits to demonstration and popularization of triplet-triplet annihilation photochemical reactions that can be driven by traditional visible-light upconversion (TTA UC) in solution using continuous-wave (cw) irradiation. UV excitation can be damaging and induce side reactions, lasers represented an important advance because multi-photon hence visible or even near-IR light is usually preferable. Thus, processes became more widely amenable,[4] requiring neither photochemistry currently faces two divergent challenges, namely the pulsed lasers nor time-resolved spectrometers. Technological desire to perform ever more thermodynamically demanding advances leading to the availability of high-power LEDs with reactions with increasingly lower photon energies. The pooling of outputs covering the entire visible spectral range further two low-energy photons can address both challenges simultaneously, contributed to making multi-photon chemistry applicable in the and whilst multi-photon spectroscopy is well established, synthetic synthetic laboratory. The fact that such processes are now photoredox chemistry has only recently started to exploit multi- exploitable for preparative-scale conversions is remarkable and photon processes on the preparative scale. Herein, we have a represents a breakthrough in photochemistry. critical look at currently developed reactions and mechanistic The field of multi-photon excitation-based photoredox catalysis concepts, discuss pertinent experimental methods, and provide an is still very young, with studies reporting preparative-scale outlook into possible future developments of this rapidly emerging reactions only since 2014. Whilst the synthetic utility for area. thermodynamically challenging reactions seems undisputable, mechanistic aspects have led to several controversial discussions. The field would likely benefit from stronger interaction between synthetically oriented organic chemists on 1. Introduction the one hand and spectroscopists as well as physical-inorganic chemists on the other hand to tackle synthetic as well as Photoredox catalysis is an old concept that has been developed mechanistic challenges for new photochemical systems. In this [1] to a remarkable level of sophistication over the past decade. spirit and working near the interface of these sub-disciplines, we Now, some inherent thermodynamic limitations of traditional prepared this article. We provide an overview of the different visible-light irradiation strategies are becoming increasingly relevant concepts of multi-photon excitation and consider their evident. Two key factors govern these thermodynamic thermodynamic and kinetic particularities based on restrictions, namely the limited energy of visible photons (up to spectroscopic studies performed with laser techniques. 300 kJ/mol), and the excitation energy loss on the way from light Furthermore, we discuss their synthetic applications and absorption by the photosensitizer till substrate activation by the important mechanistic aspects, which seem crucial for allowing reacting catalyst species. As a rule of thumb, that energy loss further rational progress in this field. High-end as well as more amounts to at least 25 % (more typically about 50%) compared straightforward techniques for obtaining fundamental insight are to the initial photon energy, which is due to the accumulated outlined, and finally we contemplate possible future directions for energy losses caused by internal conversion, intersystem research on multi-photon excitation-based photoredox catalysis. crossing, and oxidation or reduction of the catalyst itself. Hence, We hope that this article will provide some useful guidance in the range of redox potentials in which robust visible-light identifying challenges, pitfalls, and opportunities in this emerging absorbing photocatalysts are operational is rather narrow. field. Increasing the photon energy is often not an option because many organic substrates directly absorb UV light, which can cause photo-damage and undesired side reactions. A more 2. Consecutive photo-excitation in the viable alternative is the development of tailor-made photoredox simplest case catalysts with aggressive redox properties, but this requires significant efforts and can jeopardize the stability of the All currently known preparative-scale photoredox reactions photocatalysts.[2] To drive thermodynamically increasingly relying on multi-photon excitation operate based on the challenging reactions with visible light and well-established consecutive (rather than simultaneous) absorption of photons. photocatalysts, the synthetically oriented photoredox community An overview of the photon densities required for consecutive therefore recently started to exploit multi-photon excitation. By (stepwise) two-photon absorption processes in the context of this strategy, the thermodynamic hurdles are overcome by photochromic reactions has been given recently,[5] and these combining the energy of two (or more) photons per catalytic light power considerations hold also true for the processes turnover. discussed herein. However, much higher excitation densities Spectroscopists have explored multi-photon processes in than provided by laboratory-fit LEDs and cw-lasers would be considerable depth, often with pulsed lasers that provided very required for simultaneous two-photon absorption to be high excitation densities, and short-lived photoproducts were efficient.[6] The sequence of ground-state absorption followed by 2 REVIEW excited state absorption events (upper and right part of Fig. 1) pulsed Nd:YAG laser that provided high excitation densities in therefore represents the conceptually most straightforward way short periods of time. to access highly energetic species. Such sequences have been explored extensively in the field of photon upconversion with Table 1. Reactions following the mechanism in Fig. 1 along with employed [7] lanthanide doped oxide or halide materials. Trivalent photocatalysts. lanthanides in these materials often exhibit multiple very long- lived f-f excited states, and therefore sizeable excited-state populations can build up even at moderate excitation powers, and this can be exploited for the conversion of near-infrared into visible or UV light. However, in molecular compounds in fluid solution, the lifetimes of electronically excited states are typically much shorter, and consequently the deceptively simple concept in Fig. 1 becomes tricky to realize for efficient preparative-scale chemical conversions. a ref. [9], b ref. [12], c ref. [13]. Figure 1. Simplified reaction via higher excited (triplet) states. PC = photoactive compound; the asterisk denotes an electronically excited state. Aiming to exploit the mechanism of Fig. 1 under continuous- wave (non-pulsed) irradiation conditions, we discovered that the [Ir(sppy) ] complex (Table 1, entry 2) catalyzes the formation of Photoionization studies with pulsed lasers provide quantitative 3 hydrated electrons in the presence of triethanolamine (TEAO) or insight into the sequential photon absorption implied in the ascorbate.[12] A diode laser (447 nm) sufficed for the 50 mg scale mechanism of Fig. 1. For instance with xanthone as photoactive photoreduction of 4-(trifluoromethyl)benzoate to the compound (PC), a first UV photon almost quantitatively corresponding difluoromethyl compound, as well as for the produces the lowest triplet (T ) state (*PC) with a lifetime () of 1 decomposition of a benzylammonium cation. Despite the ca. 20 µs in a methanol/water mixture, which can be further importance for pharma industry, only few good methods for the excited to a higher triplet (T ) by a secondary UV photon.[8] n selective activation of CF groups seem to be known yet,[14] and Ionization from the T state occurs with a quantum yield () of 3 n the structural motif of the benzylammonium cation is present in only 0.008 due to very rapid relaxation of that T state, yet the n many widely-used quaternary ammonium compounds,[15] which formation of hydrated electrons (e -) and oxidized xanthone aq need to be degraded in wastewater. Two-pulse two-color laser (PC+) was clearly detectable. However, photoredox catalysis flash photolysis provided unambiguous evidence for the was not performed with this system; the need for UV excitation mechanism in Fig. 1: An initial 430-nm pulse populates the and the low ionization quantum yield would represent significant 3MLCT excited state of [Ir(sppy)] with a lifetime of 1.6 µs, disadvantages. 3 exhibiting strong excited-state absorption between 460 and 570 The full catalytic cycle of Fig. 1 was completed in a system nm. In this wavelength range, the ground state does not absorb comprised of 3-aminoperylene (PerNH ) as photocatalyst and 2 hence a 532-nm secondary laser pulse selectively promotes ascorbate monoanion (HAsc-) as sacrificial electron donor, 3MLCT-excited [Ir(sppy) ] to a higher triplet excited state, from enabling the green-light induced decomposition of chloroacetate