Photochemistry of Ru(II) Triazole Complexes with 6-Membered Chelate Ligands: Detection and Reactivity of Ligand-Loss Intermediates

Photochemical ligand release from metal complexes may be exploited in the development of novel photoactivated chemotherapy agents for the treatment of cancer and other diseases. Highly intriguing photochemical behavior is reported for two ruthenium(II) complexes bearing conformationally flexible 1,2,3-triazole-based ligands incorporating a methylene spacer to form 6-membered chelate rings. [Ru(bpy)2(pictz)]2+ (1) and [Ru(bpy)2(btzm)]2+ (2) (bpy = 2,2′-bipyridyl; pictz = 1-(picolyl)-4-phenyl-1,2,3-triazole; btzm = bis(4-phenyl-1,2,3-triazol-4-yl)methane) exhibit coordination by the triazole ring through the less basic N2 atom as a consequence of chelation and readily undergo photochemical release of the pictz and btzm ligands (ϕ = 0.079 and 0.091, respectively) in acetonitrile solution to form cis-[Ru(bpy)2(NCMe)2]2+ (3) in both cases. Ligand-loss intermediates of the form [Ru(bpy)2(κ1-pictz or κ1-btzm)(NCCD3)]2+ are detected by 1H NMR spectroscopy and mass spectrometry. Photolysis of 1 yields three ligand-loss intermediates with monodentate pictz ligands, two of which form through simple decoordination of either the pyridine or triazole donor with subsequent solvent coordination (4-tz(N2) and 4-py, respectively). The third intermediate, shown to be able to form photochemically directly from 1, arises through linkage isomerism in which the monodentate pictz ligand is coordinated by the triazole N3 atom (4-tz(N3)) with a comparable ligand-loss intermediate with an N3-bound κ1-btzm ligand also observed for 2.


NMR spectra for ligands and complexes S2
Photochemical studies S6

Stability studies of intermediate photoproducts S11
In-situ real-time kinetic modelling of photochemical reactivity by NMR S12

S2
NMR spectra for ligands and complexes.Experimental data fitted to kinetic model using DynaFit 4.  The remaining scenarios (see below, S is interchangeable with T) are difficult to distinguish.However, it can be noted that in any case there appear to be only two major intermediates for the last step of this photoreaction, not three (as conversion of the third intermediate to P appear to be much slower than for the other two)."Last step intermediates" therefore here are C (i.e., molecules with characteristic signals marked with circles) and either S or T. Notably, C is produced both directly (from 1), and indirectly (from either S or T).Interestingly, the rate constants appear to be very similar overall regardless of the S vs T assignment, with initial conversions slower than the final stem conversions.

Figure S9 .
Figure S9.Normalised spectra of output from the 446 nm LED (left) and 23 W fluorescent lamp (right) used in photolysis experiments.

Figure S11. 1 H
Figure S11. 1 H NMR spectra of 1 (top) and 2 (bottom) in d3-acetonitrile.Samples left in the dark to test thermal stability.

Figure S14 .
Figure S14. 1 H-1 H COSY spectrum of the region for the methylene protons of 2 after partial photolysis in d3-acetonitrile showing correlation of signals for diasteretopic CH2 protons of proposed ligand-loss intermediate 5-tz (N3) .

Figure S15 1 H
Figure S15 1 H NMR spectra of a partially photolyzed sample of 1 showing changes in upon storing in the dark for 1, 4-tz (N3) (), 4-py and 4-tz (N2) ( & ), and free pictz ligand ( †) (bottom: after 5 minutes photolysis; middle: after 1½ hours in the dark; top: sample left overnight in the dark; * solvent impurity).The light source used for photolysis was a 23 W domestic fluorescent light bulb (Hg emission lines).

Figure S16 .
Figure S16.Change in concentrations for 1, 4-tz (N3) , combined 4-py & 4-tz (N2) and free pictz ligand for a partially photolyzed NMR sample of 1. Sample was left in the dark in the bore of the spectrometer with spectra recorded at 20 minutes intervals.The light source used for photolysis was a 23 W domestic fluorescent light bulb (Hg emission lines).

(
Rates x 10 -3 % s -1 ) Irrespective of the position of S and T in the initial model, models with Direct+Indirect formation of C (4-tz (N3) ) from the initial starting material 1 (ie, Models 1 and 2 in both tables above) fit the experimental data better than models with Indirect Only formation of C from T/S, or Direct Only models.

Figure S19. 1 H
Figure S19. 1 H NMR spectra recorded during illumination of complex 2 ( ex = 459 nm).The sample was irradiated with a 2 s pulse prior to each spectral acquisition.Selected spectra taken after successive 20 s cumulative irradiation periods with signals used for kinetic analyses highlighted (starting material 2 (labelled "SM"), photoproduct 3 (labelled "P") and intermediate 5 (labelled "I")).

Figure S20 .
Figure S20.Kinetics of complex 2 photolysis studied by in situ photo-NMR ( ex = 459 nm).Experimental data fitted to kinetic model using DynaFit 4.