Olefin metathesis has become one of the most widely used techniques for C-C double bond formation. The intensive progress in this field was recognized more than a decade ago with the award of the 2005 Nobel Prize in Chemistry to Yves Chauvin, Richard Schrock and Robert H. Grubbs , , . In general, olefin metathesis is mediated by a carbene exchange between a metal carbene and an alkene. The basic mechanism of the reaction allows a variety of metathesis processes, such as cross-metathesis (CM), ring-closing metathesis (RCM), ring-opening metathesis polymerization (ROMP) and acyclic diene metathesis polymerization (ADMET) (Scheme 1). Over the years there has been considerable progress in the development of ruthenium olefin metathesis catalysts, especially based on Grubbs’ 2nd generation catalyst (Scheme 1c) , , . The introduction of N-heterocyclic carbene ligands (NHC) and chelating benzylidene ligands instead of the reactive phosphine ligands resulted in increased stability of Ru-complexes and higher catalyst activity in metathesis reactions , , .
The field of latent olefin metathesis catalysis has seen a prominent growth during the last decade , , , , , , , , , , , , , partially driven by the need to achieve controlled ROMP reactions of the more active and useful monomers, such as norbornene and dicyclopentadiene .
One of the leading families of thermally latent catalysts is the sulfur-chelated ruthenium benzylidenes , , . These complexes share a cis-dichloro configuration around the metal center, with the chelated carbene in the apical position of a slightly distorted square-pyramidal structure. Furthermore, sulfur binds ruthenium stronger than oxygen, thus it was expected that these types of complexes would show enhanced latency when compared to the Hoveyda-Grubbs catalyst. Hence, in agreement with the earlier studies of Van der Schaaf et al.  and Ung et al. , which showed that the cis-dichloro arrangement caused a decrease in activity, the S-chelated benzylidene precatalysts exhibited complete latency towards RCM at room temperature and were active only at high temperatures , . Similar latent behavior was observed in ROMP reactions: the particularly slow initiation of the S-chelated catalysts coupled with the very fast propagation, provided ROMP derived polymers with relatively low polydispersity, where the size of the polymer was controlled by the turnover number and not by the initiator to monomer ratio . Recognizing the importance of the S-chelated influence in cis/trans isomerization and the activity of Ru precatalysts, a series of studies were performed to explore the boundaries of latency in this family of complexes , , , , , , , , .
Latent olefin metathesis catalysts activated by light are also being widely developed , , , , , especially for photo-induced catalysis ,  and industrial applications , . The seminal discovery that UV light can activate the sulfur chelated catalysts  created a new method for light controlled olefin metathesis reactions. In this article, we discuss in detail (i) the photo-activation process of sulfur-chelated ruthenium complexes; (ii) how the coupling of two orthogonal photochemical reactions (photoremovable protecting group PPG cleavage 254 nm/RCM 350 nm), enabled high regioselectivity towards five- or six-membered-ring frameworks by flipping the order of the reactions ; and (iii) a PPG insertion in a light activated S-chelated ruthenium catalyst that opened a doorway for stereolithographic applications and added a new tool for complex light-guided chemical processes .
Photo-switchable ruthenium metathesis
Initially, it was envisioned that photodissociation of the sulfur-ruthenium bond  could be utilized to activate the latent catalysts. For this purpose several S-chelated ruthenium complexes were investigated, including some with aromatic substituents on the sulfur atom that absorb light at the UV-A region (Scheme 2). In all cases, the cis-dichloro isomer was obtained as thermodynamic product after 24 h reflux in DCM. Alternatively, lowering the reaction temperature, time and solvent polarity (from DCM to benzene) afforded the trans-dichloro isomers in high yields , , , , , , , , . Notably, the length of S-Ru bond in the trans isomer is quite longer (2.48 Å) than its cis counterpart (2.34 Å), possibly due to the trans influence of the strong sigma donating NHC ligand , , .
UV irradiation of cis complexes 2–4 in DCM-d2 afforded approximately 30% yield of the trans complexes (Scheme 3a). DFT calculations suggested a possible mechanism for this transformation (Scheme 3b), which was in good agreement with the previous report by Barbasiewicz and Grela on the quinoxaline chelated system . Thus, the configuration change could be explained by a photodissociation of the S-Ru bond, followed by formation of a trans-dichloro 14e ruthenium intermediate species. Subsequent ligation of the sulfur to the available site on ruthenium produces then the trans-dichloro 16e complex .
The relationship between the cis-trans ratio and olefin metathesis activity was illustrated by a benchmark ring-closing metathesis (RCM) reaction, using diethyl diallylmalonate (DEDAM) as a substrate and catalysts 3 or 4 (Table 1) . As shown in Table 1, the higher the amount of trans isomer generated in the solution, the greater conversion could be achieved after 1 day at room temperature. Conversely, when the catalyst and substrate solutions were not irradiated (only cis isomer) no sign of RCM could be detected.
Given that the cis isomer is more thermodynamically stable, heating can rapidly restore the equilibrium after irradiation; effectively converting almost all the active trans isomer to its inactive counterpart. Thus, irradiation of the mixture can generate the active complex, while short pulses of heating can deactivate it. An interesting switchable process was then planned by judicious use of light and heat. The activation cycle, UV irradiation of a solution of 4cis and DEDAM for 15 min at room temperature, was followed by 5 min heating to 80°C. As expected, UV irradiation promoted RCM, while heating stopped the reaction because it quickly depleted the active isomer. Notably, a 5 min heating period at 80°C without UV irradiation produced a negligible conversion, but was enough to remove all visible signs (NMR) of the trans isomer from solution. Further UV irradiation formed new 4trans species and resumed reaction conversion, while another round of short heating stopped it. This protocol allowed for precise quantities of DEDAM to be ring-closed at every activation/deactivation step as shown in Fig. 1. It is quite particular of this system that the activation is done by light and the deactivation by heat. Naturally, prolonged periods of heating constantly generate trace amounts of active catalyst which in the presence of substrate are consumed and continue to promote metathesis reactions, as observed in the thermally latent catalysts previously mentioned.
Chromatic orthogonality in olefin metathesis
The term chromatic orthogonality was coined by Bochet ,  to describe the selective photoreaction of a certain chromophoric functional group when irradiated at a specific wavelength in the presence of other chromophoric groups which remain intact. Despite the importance of such control in organic synthesis, most studies were focused only in protecting group chemistry , , , , . For example, two different photolabile protecting groups (PPGs), 2-nitroveratryl (Scheme 4, marked in red) and dimethoxy benzyl (marked in blue) were selectively cleaved in response to a specific wavelength (420 and 254 nm, respectively) (Scheme 4). Furthermore, the selectivity is preserved even when the PPGs are both in the same molecule due to lack of through space energy transfer during the process , . Recently, Hansen et al.  reviewed the requirements for dynamic systems involving wavelength-selective deprotection, focusing on the choice and optimization of PPGs, synthetic methods for their introduction and strategies for combining multiple PPGs into one system. Accordingly, in an ideal situation, there should be no spectral overlap between the PPGs in order to achieve chromatic orthogonality.
Sulfur-chelated ruthenium complexes are inert to photoisomerization at 254 nm. This observation led to the idea to combine photoactivation of RCM with an orthogonal photo-deprotection reaction . Based on Brook’s discovery that supersilyl groups can be photochemically cleaved at 254 nm , compound 6 was synthesized. The supersilyl protecting group in 6 was fully cleaved after 20 min irradiation at 254 nm in DCM with methanol to afford 5. Moreover, irradiation of 6 at 350 nm for 1 h did not cleave the protecting group, an essential requisite for the orthogonal system. However, irradiation at 350 nm in the presence of catalyst 4cis promoted the ring-closing metathesis of 6 to produce cyclopentene 7. These two described reactions are light-induced orthogonal protecting group deprotection/precatalyst activation and by their combination a commutative chromatic orthogonal cycle could be closed by irradiating the mixtures with both wavelengths to make the cyclopentenol final product (8) (Scheme 5).
Controlling reactions to achieve one of many possible chemical directions, i.e. chemo-, regio-, and stereoselectivity, is one of the greatest challenges in catalysis , . The most common approach to direct a reaction towards a specific direction is based on the introduction of a directing group either on the reagent/catalyst or on the substrate itself , , , . However, a more intriguing approach to affect reaction pathways could be based on “information” introduced to the studied reaction by an external stimulus. Light is the ideal external control element for in situ chemical manipulation, partly because its wavelength and intensity may be precisely regulated to achieve enhanced selectivity, and also because it can be used in the presence of a great variety of chemical functions without altering their integrity. These advantages could be also implemented by setting out an olefin-metathesis catalytic cycle in which the order of irradiation with light of different wavelengths could be used to guide reactions to specific direction (vide infra).
Inspired by the work of Schmidt et al. on the synthesis of dihydrofuranes and dihydropyranes through control of ring size selectivity by RCM reactions of triene derivatives , , , , , , , , two representative triene substrates, 9 and 10 were studied in RCM reactions promoted by the S-chelated photoactive ruthenium catalysts (Scheme 6). Each of the substrates reacted in the present of precatalysts, 3cis or 4cis, following irradiation at 350 nm to afford five- and six membered heterocycles (11a/12a and 11b/12b, respectively). As shown in Scheme 6, when 9 was photo-irradiated in the presence of the catalyst, dihydropyran 11b was obtained in favor of dihydrofuran 11a. In contrast, the influence of a sterically demanding supersilyl protecting group in compound 10 resulted in a reversal of RCM regioselectivity, i.e. affording in favor the five-membered ring product, 12a was the major product. These results are in complete accordance with Schmidt’s previous findings , .
Thus, by following the chromatic orthogonality principles a novel regioselective synthetic pathway towards dihydrofuranes or dihydropyranes could be carried out using only the order of wavelength irradiation as the guiding force (Scheme 7). Pathway A started by irradiation of 10 at 350 nm to afford mainly five membered RCM products, followed by irradiation at 254 nm to produce 11a as the major product. Alternatively, pathway B began with the irradiation of 10 at 254 nm to afford alcohol 9 and then irradiation at 350 nm delivered the the six-membered ring RCM product 11b with good selectivity (6:1). This is a unique example where chromatic orthogonality is used in a light activated catalytic sequence together with a photolabile protecting group to achieve a remarkable regioselective synthesis.
A smart catalyst equipped with a chromatic orthogonal self-destructive function
Light-activated olefin metathesis, especially in ring-closing metathesis polymerization (ROMP) can offer a broad range of applications for new materials and surface modification , , , . For example, in thin film generation, the use of dormant catalysts may allow better control over patterning processes or device manufacturing if it would be possible to switch “on” the pre-catalyst by a certain wavelength light irradiation and then to switch it “off” permanently under irradiation at a different wavelength.
Taking into account that functional groups on the NHC ligand of ruthenium benzylidenes have a strong effect on catalyst activity , , , ,  and having observed that the photodeprotection of supersilyl groups under certain conditions promoted the decomposition of Grubbs type catalysts , we set out to examine whether adding the supersilyl directly on the NHC could provide an orthogonal chromatic self-destruct mechanism. For this purpose a new sulfur chelated ruthenium catalyst bearing two supersilyl protecting groups at the phenol side arms of the NHC ligand was prepared . When subjected to irradiation at 350 nm, complex (13cis) was photoisomerized to the catalytically active trans-dichloro form in about 30% yield (Fig. 2), consistent with the behavior of other S-chelated precatalysts (vide supra).
Irradiation of a DCM solution containing diethyl diallylmalonate (DEDAM) and 13cis (5%-mol) at 254 nm (UV-C) and 350 nm (UV-A) expectedly afforded different results (monitored by 1H NMR). Irradiation with UV-A light resulted with 95% conversion of DEDAM to the cyclized product, whereas irradiation with UV-C light did not promote ring closing metathesis (RCM), and only unreacted DEDAM could be seen (Scheme 8a). Interestingly, further irradiation with UV-A (after the sample had been treated by UV-C) did not promote RCM, suggesting that all 13 had been depleted by UV-C irradiation. A control experiment was also carried-out using the following irradiation sequence: a solution containing DEDAM and 13cis (5%-mol) in DCM was irradiated by UV-A light until 50% conversion was reached. Then the solution was irradiated by UV-C, and immediately after with UV-A; however no further progress in RCM could be detected (Scheme 8b). In contrast, when catalyst 3cis was used, the irradiation with UV-C did not fully decompose the catalyst and further irradiation with UV-A did bring about RCM of DEDAM (Scheme 8c). These experimental results suggested that the photocleavage of the supersilyl protecting groups from precatalyst 13cis resulted in catalyst deterioration and decomposition.
The activation/deactivation protocol could also be used for other types of olefin metathesis reactions, especially in applications dealing with polymerizations . Thus, the photoinduced ring opening metathesis polymerizations (PROMP) , ,  of exo-dimethyl-5-norbornene-2,3-dicarboxylate norbornene (14) and cyclooctene (15) were studied by using the appropriate UV irradiation sequence. In both examples, irradiation with UV-A light of the monomers with 13cis in CH2Cl2 afforded complete polymerization (Scheme 9). The afforded polymers had high molecular weights and low polydispersities. However, if UV-C light (254 nm) was applied in the first step, the catalyst was permanently inhibited and no polymer was obtained.
As a final proof of concept of the chromatic self-destruct function, the catalyst 13cis performance was examined on a polymerizable monomer layer to simulate a 3-D printing process. Hence, monomer 14 and the latent catalyst 13cis were dissolved in a minimal amount of CH2Cl2 and then solution was placed in two metallic molds; one of which was irradiated at 350 nm for 15 min (mold 1) and the other one (mold 2) was irradiated with a sequence of 254 nm for 20 min and 350 nm for 15 min. Polymer samples were removed from the molds, washed with methanol and analyzed by 1H NMR and GPC. Figure 3 shows the striking visual contrast between the two molds demonstrating a fully polymerized monomer and an almost completely inhibited reaction. Interestingly, a small amount of polymer was still seen along the edges of mold 2. It seems that irradiation close to the edges with 254 nm was not enough to completely decompose the catalyst in the time allotted.
To confirm the complete inhibition of catalyst, the reactions were repeated in quartz tubes with 1 M monomer in DCM and 0.036%-mol of catalyst (Fig. 4). The polymer was observed only in tube number 2, which was irradiated at 350 nm for 15 min, after addition of excess methanol (monomer is soluble in methanol while the polymer is not). In contrast, polymerization was not observed in tube number 1, which was first irradiated at 254 nm for 20 min and only then at 350 nm for 15 min. 1H-NMR analysis of both samples (after evaporation of solvent and dissolving in CDCl3) confirmed that the reaction mixture in tube number 1 consisted only of unreacted 14 whereas tube 2 contained both monomer and resultant polymer (Fig. 4).
In this review article we have summarized the photoactivity and properties of latent sulfur-chelated ruthenium olefin metathesis precatalysts. Photoactivation was induced by photoisomerization of the more thermodynamically stable cis-dichloro isomer to the trans-dichloro form. The ability of these dormant catalysts to be active upon UV-A light provided an opportunity to design a model system that can be chromatically orthogonal with photochemical reactions that are promoted by UV-C. Thus, orthogonal photoactivated RCM and photo-removal of protecting groups was achieved and the regioselective pathway of reactions could be changed by altering the irradiation sequence. Finally, the addition of a supersilyl protecting group to the NHC portion of the catalyst led to an orthogonal chromatic kill switch, allowing for a permanent deactivation of olefin metathesis reactions. These proof-of-concept examples open a doorway for stereolithographic applications and adds new tools for complex light guided chemical processes.
We gratefully acknowledge financial support by The Israel Science Foundation (ISF), the Magnet Program of the Office of the Chief Scientist of The Ministry of Industry, Trade and Labor and The Open University internal funds.
E. Tzur, G. Lemcoff in Handbook of Metathesis Vol. 3: Polymer Synthesis. R. H. Grubbs, E. Khorsavi (Eds.), Wiley-VCH, Chap. 12 (2012). Google Scholar
E. Alessio, G. Mestroni, G. Nardin, W. M. Attia, M. Calligaris, G. Sava, S. Zorzet. Inorg. Chem. 27, 4099 (1998). Google Scholar
For extensive discussion on trans influence in chelating alkylidene ligands see ref. 28. Google Scholar
For review on phtotoremovable protecting groups in chemistry and biology see: P. Klan, T. Solomek, C. G. Bochet, A. Blanc, R. Givens, M. Rubina, V. Popik, A. Kostikov, J. Wirz. Chem. Rev. 113, 119 (2013). CrossrefGoogle Scholar
A. Blanc, C. G. Bochet. Org. Lett. 9, 2649 (2009). Google Scholar
C. G. Bochet. Synlett 13, 2268 (2004). Google Scholar
C. G. Frost, P. Marce, P. M. Liu. Organomet. Chem. 40, 54 (2016). Google Scholar
B. Schmidt, S. Nave. Chem Commun. 2489 (2006). Google Scholar
B. Schmidt. Synlett 2004, 1541 (2004). Google Scholar
H. Ito, C. G. Willson. in Polymers in Electronics. T. Davidson (Ed.), ACS Symposium Series, 242; ACS Publications: Washington, DC, pp 11–23 (1984). Google Scholar
M. Süßner, H. Plenio. Chem. Commun. 5417 (2005). Google Scholar
S. Manzini, C. A. Urbina-Blanco, D. J. Nelson, A. Poater, T. Lebl, S. Meiries, A. M. Z Slawin, L. Falivene, L. Cavallo, S. P. Nolan. J. Organomet. Chem. 2015, 780, 43 (2015). Google Scholar
About the article
Published Online: 2017-03-17
Published in Print: 2017-06-27
Citation Information: Pure and Applied Chemistry, Volume 89, Issue 6, Pages 829–840, ISSN (Online) 1365-3075, ISSN (Print) 0033-4545, DOI: https://doi.org/10.1515/pac-2016-1221.http://creativecommons.org/licenses/by-nc-nd/4.0/.