Scope for Late-Stage Diversification through the remote meta-C─H Activation

remote C–H functionalizations was applicable to structurally complex electrophiles (Scheme 66).

Late-stage diversification of purine bases with BODIPY fluorescence labels was accomplished by carboxylate-phosphine ruthenium catalysis (187a and 187b). Notably, electrophiles bearing amino acids were smoothly transformed to the corresponding products 187c–187i with high levels of chemo-selectivity without any evidence for racemization. It was highlighted that reactive unprotected hydroxyl groups in serine (187f) and tyrosine (187i) as well as free NH-indole in tryptophan (187h) were fully tolerated. In addition, more structurally complex peptides underwent the desired chemical ligation to form products 187j and 187k, featuring among others sensitive methionine. The synergistic ruthenium(II) catalyst verified also fully compatible with triglycerides derived from saturated and unsaturated fatty acids (187l–187o) and vitamin

D-α-tocopherol (187p). Particularly, the chemo-selectivity of the meta-C–H transformation in the presence of unsaturated fatty acids (187n and 187o) is noteworthy, since they are simply disposed to olefinic and allylic functionalizations. Remarkably, the late-stage modification of marketed drugs was accomplished, including transformations of neuroprotective agent gastrodin (187q–

187s) and anti-inflammatory salicin (187t). The ruthenium catalysis was not restricted to benzylic electrophiles, but also synthetically useful monosaccharide bromoesters afforded the desired meta-alkylated products 187u–187w with high catalytic efficacy. It is noteworthy that fully unprotected OH-free monosaccharides (187t) proved to be compatible for the first time in ruthenium catalysis. Notably, purine-uridine hybrids 187x and 187y were obtained by the synergistic catalysis via catalytic nucleoside ligation.

3 Results and Discussion

Scheme 66: Late-stage diversification of structurally complex drugs and natural product molecules by remote meta-C–H functionalization.

3.4 Late-Stage Diversification by Selectivity Switch in meta-C–H Activation

Scheme 66 (cont.): Late-stage diversification of structurally complex drugs and natural product molecules by remote meta-C–H functionalization.

3 Results and Discussion

Scheme 66 (cont.): Late-stage diversification of structurally complex drugs and natural product molecules by remote meta-C–H functionalization.

3.4 Late-Stage Diversification by Selectivity Switch in meta-C–H Activation Although the synergistic ruthenium catalysis proved to be robust and versatile, low conversion of pyrazoles (143o) and azobenzenes (143t) was noted (Scheme 67). The reactions of oxazolinylbenzenes with different primary benzyl chlorides delivered the corresponding product 143p–143s in low yields. Moreover, fully unreactive arenes, such as O-methyloximes, dimethylpyrazoles, N-pyrimidylanilines, benzodiazepines, 2-pyridylpyridone, and others, as well as ineffective alkyl chlorides or bromides are listed in Scheme 67.

Scheme 67: Unsuccessful results and ineffective substrates in the synergistic ruthenium-phosphine catalysis.

3 Results and Discussion

3.4.5 Mechanistic Studies

Since the carboxylate-phosphine ruthenium catalysis effectively transformed several electrophiles, competition experiments were conducted to examine their reactivities (Scheme 68).

The results showed that primary benzyl chloride 142b is less active than bromoesters 84a, 140a, and 140k, whereas no reactivity was observed in case of bromocycloheptane (136h).

Scheme 68: Intermolecular competition experiments of electrophiles under the carboxylate-phosphine ruthenium catalysis.

Moreover, an intermolecular competition experiment of arenes 68a and 68b highlighted electron-rich arenes to be more efficiently converted (Scheme 69)

3.4 Late-Stage Diversification by Selectivity Switch in meta-C–H Activation

Scheme 69: Intermolecular competition experiment of pyridines 68a and 68b.

To understand the cleavage of C–X bond, experiments with radical scavengers were investigated (Scheme 70). The typically used radical scavenger TEMPO fully inhibited the synergistic C–H transformations. The isolated TEMPO adduct 191 strongly indicated the homolytic C–X bond cleavage. While the addition of BHT did not have any influence on the catatytic potential, the reaction with 1,1-diphenylethylene significantly reduced the formation of the meta-benzylated product 143b. Moreover, the adduct derived from 1,1-diphenylethylene and benzyl chloride was detected by GC-MS spectrometry. These findings suggested a radical mechanism through homolytic C–X bond cleavage.

Scheme 70: meta-Benzylation in the presence of radical scavengers.

Afterwards, isotopically labelled substrates [D]2-68b and [D]3-68b were employed in the synergistic catalysis (Scheme 71). The substrates were efficiently converted into the corresponding product [D]n-143d. The deuteration degree of product [D]n-143d was measured by

1H-NMR spectroscopy, where H/D scrambling at ortho position was observed. Owing to a trace amount of H2O from K2CO3, it could not indicate proton source in protodemetalation process.

However, it was suggestive of a reversible C–H metalation process.

3 Results and Discussion

Scheme 71: Ruthenium-catalyzed meta-benzylation of isotopically labeled substrates [D]2- and [D]3-68b.

To delineate the mechanism of the synergistic C–H functionalization, the series of well-defined ruthenium intermediates were prepared (Scheme 72). First, ¹H-NMR spectroscopic studies revealed the p-cymene dissociation from the ruthenium precatalyst, which suggested p-cymene-coordinated ruthenacycle was not involved in the catalytic cycle. Second, the obtained single crystals from mixture of pyridine 68b, [Ru(OAc)2(p-cymene)] (181), and PPh3 were analyzed by X-ray crystallography, providing the structure of ruthenacycle trans-192a with two phosphines (Scheme 72a). Due to two phosphine equivalents on complex trans-192a, an additional equivalent of phosphine was added into the complex mixture, affording 59% of complex trans-192a (Scheme 72b). Moreover, pyridine 68d gave under the same conditions a 1:17 mixture of cis- and trans-ruthenacycle 192b. The alternative protocol with two equivalents of phosphine ligands delivered a 1:1 mixture of cis- and trans-ruthenacycle 192a (Scheme 72c). In the case of the bidentate phosphine ligand DPEPhos, monocyclometalated complex 193 was obtained and confirmed by X-ray crystal structure analysis (Scheme 72d).

3.4 Late-Stage Diversification by Selectivity Switch in meta-C–H Activation

Scheme 72: Preparation of well-defined ruthenacycles 192 and 193.

In addition, the carboxylate-phosphine ruthenium complex trans-192a could be derived from cationic monocyclometalated complex 98 with two equivalents of phosphine (Scheme 73a). In contrast, the reaction with one equivalent of phosphine smoothly delivered ruthenacycle 194 (Scheme 73b). The structure of complex 194 was established by X-ray crystallography.

3 Results and Discussion

Scheme 73: Ligand modification of cyclometalated ruthenium complex 98.

Having a series of well-defined ruthenacycles in hand, the isomerization of complex 192a was examined by ³¹P{¹H}-NMR spectroscopy (Scheme 74). The solution of complex trans-192a in THF-d8 was heated at 60 °C, affording a 0.3:1.0 mixture of cis- and trans-192a. This finding indicated that one of the phosphine ligands is simply labile in the ruthenium complex. These results are in good agreement with the Ru–P bond lengths of well-defined ruthenium complexes.

According to X-ray crystallographic data, ruthenacycle trans-192a has Ru–P bond lengths of 2.3421 and 2.3355 Å, which are longer than complex 194 with a bond length of 2.2451 Å. In addition, bidentate phosphine ruthenium complex 193 has an axial Ru–P bond length of 2.2377 Å and an equatorial Ru–P bond length of 2.3263 Å. These data supported the weak coordination of phosphine ligand on ruthenium complex trans-192a.

Scheme 74: Isomerization of trans-192a by temperature.

Afterwards, the second C–H metalation of monocyclometalated complex cis-/trans-192a and pyridine 68d was investigated (Table 16). ³¹P{¹H}-NMR spectroscopic studies did not detect any construction of the biscyclometalated ruthenium complex. However, ligand exchange of ruthenacycles cis-/trans-192 with pyridine 68d was observed in the presence and absence of K2CO3, leading to the formation of ruthenium complexes cis-/trans-192b.

3.4 Late-Stage Diversification by Selectivity Switch in meta-C–H Activation Table 16: Studies on ligand exchange of ruthenacycle cis-/trans-192a and pyridine 68d.

cis-192a trans-192a cis-192b trans-192b phosphine ligands significantly reduced oxidation potential of ruthenium(II/III) complexes. In contrast, the p-cymene-coordinated ruthenium complex 195 exhibited an irreversible oxidation event at E = 0.81 V.

Figure 10: Cyclic voltammetry studies in 1,2-DCE containing 0.1 mol∙L^–1 n-Bu4NPF6, scan rate 100 mV∙s^–1.

3 Results and Discussion

Figure 10 (cont.): Cyclic voltammetry studies in 1,2-DCE containing 0.1 mol∙L^–1 n-Bu4NPF6, scan rate 100 mV∙s^–1.

Then, the catalytic efficacy of well-defined ruthenacycle 192a–194 was evaluated in the remote meta-C–H benzylations (Table 17). Complex trans-192a and 194 efficiently delivered the desired product 143d, while bidentate phosphine ruthenium complex 193 failed to give any conversion of substrate 68b.

Table 17: Remote meta-C–H benzylation catalyzed by cyclometalated ruthenium complexes.^[a]

3.4 Late-Stage Diversification by Selectivity Switch in meta-C–H Activation

Entry [Ru] 143d (%)

1 trans-192a 54 (39)

2 193 --- (---)

3 194 59 (50)

[a] Reaction conditions: 68b (0.50 mmol), 142b (1.50 mmol), [Ru] (10 mol %), KOAc (10 mol %), K2CO3 (1.0 mmol), 1,4-dioxane (2.0 mL), 60 °C, 20 h, under N2; yield of isolated products. The yield in parentheses was obtained in the absence of KOAc.

Furthermore, the stoichiometric experiments of ruthenacycle 192a with benzyl chloride 142b selectively afforded the corresponding product 143d (Scheme 75a). It is noteworthy that demetalation by the addition of 2,2’-bipyridine and acetic acid was obligatory after the transformation. On the other hand, carboxylate-phosphine ruthenium complex 194 failed to provide the desired product 143d (Scheme 75b).

Scheme 75: Stoichiometric benzylation of ruthenacycle 192a and 194.

3.4.6 Proposed Catalytic Cycle

On the basis of experimental and computational findings,^[104] a plausible catalytic cycle commences by carboxylate-assisted ortho-C–H ruthenation to generate complex trans-192a (Scheme 76). Then, single-electron transfer (SET) from the ruthenium(II) complex trans-192a to the benzyl halide 142, generates the ruthenium-(III) intermediate 196. The benzyl radical 197 attacks on the arene moiety at the position para to ruthenium, providing triplet species 198. Next, ligand-to-metal charge-transfer leads to the significantly stabilized singlet ruthenacycle 199.

3 Results and Discussion

Finally, rearomatization and ligand exchange delivers the desired meta-benzylated product 143 and regenerates ruthenium(II) complex trans-192a.

Scheme 76: Proposed catalytic cycle for ruthenium-catalyzed remote C–H benzylation.