Regioselective Hydroxylation of Unsymmetrical Ketones Using Cu, H2O2, and Imine Directing Groups via Formation of an Electrophilic Cupric Hydroperoxide Core

Herein, we describe the regioselective functionalization of unsymmetrical ketones using imine directing groups, Cu, and H2O2. The C–H hydroxylation of the substrate–ligands derived from 2-substituted benzophenones occurred exclusively at the γ-position of the unsubstituted ring due to the formation of only one imine stereoisomer. Conversely, the imines derived from 4-substituted benzophenones produced E/Z mixtures that upon reacting with Cu and H2O2 led to two γ-C–H hydroxylation products. Contrary to our initial hypothesis, the ratio of the hydroxylation products did not depend on the ratio of the E/Z isomers but on the electrophilicity of the reactive [LCuOOH]1+. A detailed mechanistic analysis suggests a fast isomerization of the imine substrate–ligand binding the CuOOH core before the rate-determining electrophilic aromatic hydroxylation. Varying the benzophenone substituents and/or introducing electron-donating and electron-withdrawing groups on the 4-position of pyridine of the directing group allowed for fine-tuning of the electrophilicity of the mononuclear [LCuOOH]1+ to reach remarkable regioselectivities (up to 91:9 favoring the hydroxylation of the electron-rich arene ring). Lastly, we performed the C–H hydroxylation of alkyl aryl ketones, and like in the unsymmetrical benzophenones, the regioselectivity of the transformations (sp3 vs sp2) could be controlled by varying the electronics of the substrate and/or the directing group.


■ INTRODUCTION
The direct functionalization of C−H bonds is a powerful tool for organic synthesis. 1,2−5 Imine directing groups are widely used due to their facile installation (i.e., condensation between amine and aldehyde or ketone) and removal, which, in selected examples, allows for their utilization as transient directing groups. 6,7The regioselectivity of these metal-promoted C−H functionalization reactions is usually based on the formation of one of the imine stereoisomers (E), which is usually accomplished by employing aldehydes and methyl ketones as substrates (Figure 1A).With the exception of the Rh-catalyzed α-C−H hydroacylation of aldimines, 8−10 the metal is proposed to coordinate to the imine N donor, which promotes the selective activation of the βand γ-C−H bonds of the most sterically hindered carbonyl substituent (i.e., functionalization occurs at substituent trans to the directing group N-substituent, see Figure 1A). 8,11,12−20 The selectivity of these Cu-promoted oxidations, which were initially developed by Schoënecker and co-workers for the selective oxidation of steroids 21 and have been applied to the total synthesis of complex molecules, 22−24 also relies on the formation of only one of the imine isomers in unsymmetrical substrates (e.g., cyclohexyl phenyl ketone 18 ), an issue that can be avoided in the hydroxylation of symmetrical substrates (e.g., symmetrical benzophenones 18 ).In this paper, we studied the regioselectivity in the Cu-directed hydroxylation of unsymmetrical ketones (Figure 1B).We initially hypothesized that the regioselectivity of these Cu-promoted C−H hydroxylation reactions would be determined by the ratio of the imine E/Z isomers of the substrate−ligand.Contrary to our initial thoughts, the regioselectivity of these transformations is dictated by the electrophilicity of the hydroxylating Cu II OOH intermediate.

■ RESULTS AND DISCUSSION
Synthesis of Imines Derived from 2-Picolylamine and Unsymmetrical Benzophenone, and Determination of Imine E/Z Isomer Ratio.−20 The Journal of Organic Chemistry bulkier group on the carbon atom is positioned trans to the nitrogen substituent (e.g., imines derived from benzaldehyde only form the E-isomer in solution, see Figure 2A).In the case of ketones containing substituents with similar size (e.g., substituted benzophenones), it was found that the ratio of E/ Z isomers is highly dependent on the position of the substituents.Imines derived from 2-substituted benzophenones only produced Z isomers while 4-substituted benzophenones produced mixtures (ca.E/Z: 60/40).Boyd and co-workers suggested that two competing effects determined the E/Z ratio of the mixtures, a stabilizing planar resonance effect between the aromatic substituent and the imine double bond and a destabilizing steric/electronic repulsion between the aromatic substituents.For 2-substituted benzophenones, it was proposed that the aromatic ring containing the substituent twists to avoid steric repulsion with the unsubstituted ring, which adopts a planar conformation with the imine double bond.For all of the imines analyzed, minor changes in the E/Z ratios were found upon changing the imine substituent or when different solvents were utilized.
In this work, the synthesis of the imine substrate−ligands derived from substituted benzophenones and 2-picolylamine was carried out in a Dean−Stark apparatus using toluene as the solvent and catalytic amounts of p-toluenesulfonic acid monohydrate (see example in Figure 3; see also the Supporting Information, SI, for details on the synthetic procedure and characterization of the imine substrate−ligands).After isolation of the imine substrates, we determined the ratio of E/Z isomers by NMR ( 1 H NMR, 13 C NMR, COSY, and NOESY, see the SI).The imine systems derived from the 2-substituted benzophenones (2-MeO, 2-Me, 2-F, 2-Cl, and 2-Br) produced only the Z isomer, in agreement with the report by Boyd and co-workers. 25or imines derived from 4-substituted benzophenones (4-MeO, 4-Me, 4-F, 4-Cl 4-Br, and 4-CF 3 ), we observed mixtures of E/Z isomers in which the E isomers were slightly more favored than the Z (ca.E/Z: 60/40), also in agreement with literature precedents. 25ydroxylation of Imine Substrate−Ligands.An example of the protocol we followed to analyze the Cu-promoted hydroxylation of unsymmetrical benzophenones is included in Figure 3 (hydroxylation of 4-fluorobenzophenone using 2picolylamine as a directing group).After synthesizing the imine substrate−ligand and determining the purity and E/Z ratio by NMR (see 4F L E and 4F L Z in Figure 3), we carried out the hydroxylation in acetone at room temperature using 1 equiv of [Cu I (CH 3 CN) 4 ](PF 6 ) and 5 equiv of 30% aqueous H 2 O 2 .NMR Figure 2. E/Z equilibrium for imines derived from benzaldehyde, alkyl aryl ketones, and substituted benzophenones. 25e Journal of Organic Chemistry analysis allowed for determining the hydroxylation yield, the mass balance of the reaction, and the ratio of hydroxylation products derived from the imine substrate−ligand E/Z isomers (see 4F PL E and 4F PL Z in Figure 3).Removal of the directing group was accomplished using aqueous 1 M HCl, and the ratio of the resulting hydroxy benzophenone products agreed with the E/Z ratio calculated before imine cleavage (e.g., 4F PL E / 4F PL Z ∼ 4F L E / 4F L Z , see Figure 3 and SI for further details).
In the hydroxylation of the substrate−ligands derived from 2substituted benzophenones, we observed selective γ sp 2 C−H hydroxylation of the unsubstituted phenyl ring ( 2X PL E / 2X PL Z , 0/100), which agreed with the presence of only one of the imine isomers in the starting substrate−ligands ( 2X L E / 2 L Z , 0/100, see

The Journal of Organic Chemistry
Figure 4A).The hydroxylation yields (from 40 to 64%) and the mass balance of the reactions (from 60 to 84%) were comparable to the results obtained in our previous publications.
The Cu-promoted hydroxylation of the imine systems derived from the 4-substituted benzophenones also led to hydroxylation products with good yields (35−70%) and good mass balance (65−81%, see Figure 4B).Unexpectedly, the ratio of the two hydroxylation products, namely, 4X PL E and 4X PL Z , did not match the E/Z ratio of the initial imine substrate−ligands.While the Cu-promoted oxidation of 4MeO L (E/Z: 67/33) and 4Me L (E/Z: 52/48) produced hydroxylation products with E/Z ratios similar to the imine substrate−ligands (e.g., 4MeO PL E / 4MeO PL Z : 62/38), we observed that 4F L (E/Z: 64/36), 4Cl L (E/Z: 62/38), 4Br L (E/Z: 62/38), and 4CF3 L (E/Z: 60/40) favored the formation of the Z hydroxylation product (e.g., 4CF3 PL E / 4CF3 PL Z : 18/82).Thus, it appeared that the regioselectivity for the hydroxylation of the 4-substituted systems was determined by the electronics of the substituent, with the electron-donating substituents favoring the functionalization of the substituted phenyl ring and the electronwithdrawing substituents favoring the functionalization of the unsubstituted phenyl ring.
Mechanistic Studies on Regioselectivity.The proposed mechanism for the regioselective Cu-directed hydroxylation of sp 2 C−H bonds is depicted in Figure 5.In our previous publications, we have proposed that the hydroxylation of aryl substrates (e.g., imine substrate−ligands derived from symmetrical benzophenones) occurred via mononuclear Cu/O 2 species. 18Our experimental and theoretical data supported the formation of a Cu II -hydroperoxo intermediate before the ratedetermining step of the reaction.The [LCu II OOH] 1+ species can be formed by adding H 2 O 2 to copper(I) or copper(II) sources or by oxidation of [LCu I ] 1+ with O 2 , in which copper(II) and H 2 O 2 (a real oxidant) are formed via superoxide disproportionation and solvent oxidation. 17For the unsymmetrical benzophenones described herein, we hypothesized that Figure 4. Cu-promoted hydroxylation of imine substrate−ligands derived from 2-substituted benzophenones (A) and 4-substituted benzophenones (B) using 2-picolylamine as a directing group and H 2 O 2 as an oxidant (see the SI for experimental details on the synthesis of substrate−ligands, hydroxylation reactions, and directing group removal).

The Journal of Organic Chemistry
for each of the mononuclear species formed before the r.d.s., an equilibrium between the E/Z imine isomers exists but the ratio of the hydroxylation products is solely determined during the rate-determining step.We have previously shown that during the r.d.s., the [LCu II OOH] 1+ species oxidizes the aromatic sp 2 C−H bonds via a concerted heterolytic O−O bond cleavage with concomitant electrophilic attack on the arene system. 18or some of the imine substrate−ligands analyzed, we observed that the addition of Cu I induced a change in the E/Z ratios (Figure 6).The change in the E/Z ratio could be analyzed by 1 H NMR due to the formation of diamagnetic [LCu I ] 1+ species in solution (see the SI for details on NMR analysis).
Addition of 1 equiv of Cu I to 4F L (E/Z: 70/30) shifted the equilibrium to [ 4F LCu I ] 1+ (E/Z: 57/43, entry 2 in Figure 6).However, the change in the ratio upon addition of Cu I could not account for the E/Z ratio of the hydroxylation product, 4F PL (E/ Z: 38/62, entry 3 in Figure 6).We also analyzed the ratio of the 4F L imine isomers upon addition of copper (Cu I or Cu II ) and removal of Cu with Na 2 EDTA (entries 4 and 5 in Figure 6), and we observed no major changes in the imine substrate−ligand E/ Z ratios, suggesting that the coordination of Cu did not lead to irreversible isomerization of the imine substrate−ligands.Hydroxylation of 4F L with Cu I and H 2 O 2 at different  The Journal of Organic Chemistry temperatures (entries 6 to 8 in Figure 5) and with different solvents (entries 9 to 11 in Figure 6) did not substantially change the E/Z hydroxylation ratios.Like in the hydroxylation of 4F L with Cu I and H 2 O 2 , the oxidations using Cu II , NMe 4 OH, and H 2 O 2 led to an E/Z ratio of 38/62 (entry 12 in Figure 6).Likewise, the hydroxylation of 4F L with Cu I and O 2 at 50 °C led to similar E/Z ratios (entry 13 in Figure 6).
Overall, coordination of Cu I to the 4-substituted substrate− ligands caused a change in the E/Z ratio, while no change in the E/Z ratio was observed in the cuprous complexes of the 2substituted analogues.However, these changes in the E/Z ratio of the imine substrate−ligands upon addition of Cu I did not match the ratios of the hydroxylation products, suggesting that the addition of H 2 O 2 triggers the formation of the electrophilic [LCu II OOH] 1+ intermediate, which will determine the regioselectivity of these reactions during the rate-determining step (see Figure 5).
Redox Potentials and Electrophilicity.In our previous publication on the Cu-directed hydroxylation of imine substrates derived from symmetrical 4,4′-disubstituted benzophenones and 2-picolylamine, we observed that the decay of the reactive [LCuOOH] 1+ species (rate-determining step) was not influenced by the substituents on the benzophenone. 18We rationalized those findings based on the electrophilicity of the reactive CuOOH core and the relative reactivity of the substituted phenyl rings (Figure 7A).When compared to benzophenone, the system derived from 4,4′-dimethoxybenzophenone produced a CuOOH core with diminished electrophilicity, which was compensated for by the higher reactivity of the arene substrate.Conversely, the system derived from 4,4′dichlorobenzophenone produced a CuOOH core with The Journal of Organic Chemistry enhanced electrophilicity, which was compensated for by the lower reactivity of the arene substrate.To provide experimental evidence for the electrophilicity of the CuOOH cores, we measured the reduction potentials of the LCu I complexes, 26−29 and we observed remarkable variations of E 1/2 (E 1/2 : −160 mV vs Fc 0/+ for 4,4′-dimethoxybenzophenone; E 1/2 : 20 mV vs Fc 0/+ for 4,4′-dichlorobenzophenone).
Based on the proposed mechanism depicted in Figure 5, the E/Z isomers of the reactive [LCu II OOH] 1+ are in fast equilibrium and the isomer that directs the electrophilic CuOOH core toward the more electron-rich ring will react faster than the isomer that directs the CuOOH core toward the electron-poor ring (Figure 7B).Hence, the hydroxylation of the imines derived from 4-substituted benzophenones with electron-donating substituents (4-MeO and 4-Me) favors the hydroxylation of the substituted ring (electron-rich arene), and the hydroxylation of the imines derived from 4-substituted benzophenones with electron-withdrawing substituents (4-F, 4-Cl, 4-Br, and 4-CF 3 ) favors the hydroxylation of the unsubstituted ring (which is also the electron-rich arene).The relative reaction rates for the hydroxylation of substituted (k X ) and unsubstituted (k H ) rings can be obtained from the

The Journal of Organic Chemistry
4X PL E / 4X PL Z ratios ( 4X PL E / 4X PL Z = k X /k H ), which allowed constructing a Hammett plot (Figure 7B).−32 Like we did for the systems derived from symmetrical 4,4′disubstituted benzophenones, we measured the E 1/2 for the 1electron oxidation of the Cu I complexes derived from 2picolylamine and unsymmetrical 4-substituted benzophenones (Figure 7B, see also SI for experimental details).When the E 1/2 values were plotted against the relative reaction rates, we obtained a linear correlation (slope: −0.013), suggesting that the electrophilicity of the CuOOH core could be benchmarked by measuring the E 1/2 values of the Cu complexes.Moreover, we envisioned that the regioselectivity of the Cu-directed hydroxylation reactions could be potentially controlled by changing the electrophilicity of the CuOOH core via substrate and/or directing group modification (see sections below).
Hydroxylation of 4,4′-Disubstituted Benzophenones.Based on the mechanistic findings described above, we hypothesized that the utilization of 4,4′-disubstituted benzophenones could enhance the regioselectivity of the Cu-directed hydroxylation reactions (Figure 8).Imine systems derived from 4,4′-disubstituted benzophenones containing electron-rich substituents on one of the arene rings (e.g., X: MeO) and electron-poor substituents on the other arene ring (e.g., Y: F, Cl, Br, CF 3 ) were synthesized, and the A/B isomers ratios were determined by NMR (Note: instead of E/Z isomers, these are denoted as A/B to avoid misunderstanding).
As we expected, the electrophilic character of the putative CuOOH intermediate favored the oxidation of the electron-rich phenyl ring.For example, 4MeO4 ′ F L (A/B: 51/49, see Figure 8A) was oxidized with Cu The relative reaction rates between the hydroxylation of the arene ring containing the MeO substituent (k MeO ) and the hydroxylation of the other arene ring (k Y , Y: Me, H, F, Cl, CF 3 ) can be calculated from the , which allowed constructing a Hammett plot (Figure 8B; Note: for the x axis, we used the difference between σ p(MeO) and σ p(Y) ).The slope obtained (ρ = −2.38) is similar to the value obtained in the Hammett plot constructed with the systems derived from 4-substituted benzophenones (see Figure 7B), suggesting the involvement of an electrophilic oxidant (CuOOH core) during the ratedetermining step.Further support for this proposal was obtained when we plotted the relative hydroxylation rates (k MeO /k Y ) against the difference between the E 1/2 values of cuprous systems derived from the 4,4′-(MeO) 2 and 4-MeO-4′-Y systems (electrophilicity plot in Figure 8B).The slope obtained (−0.0125) is analogous to the one found in the electrophilicity plot of the 4-substituted benzophenones systems (slope = −0.013,see Figure 7B), suggesting the formation of the same electrophilic species (CuOOH core) in both systems.
Hydroxylation of Unsymmetrical 4-Substituted Benzophenones with Varying Directing Groups.The Cudirected hydroxylation of 4-MeO-, 4-Cl-, and 4-CF 3 -benzophe-Figure Cu-promoted hydroxylation of systems derived from unsymmetrical 4-substituted benzophenones and varying directing groups using H 2 O 2 as an oxidant (see the SI for experimental details on the synthesis of substrate−ligands and hydroxylation reactions).Note: the regioselectivity results ( 4X PL 4R-py E / 4X PL 4R-py Z column) are color-coded according to orange (results obtained with 2-picolylamine), red (worsened regioselectivity for other DGs when compared to 2-picolylamine), or green (improved regioselectivity for other DGs when compared to 2-picolylamine).

KETONES
Encouraged by the results obtained in the regioselective hydroxylation of substituted benzophenones described above, we decided to explore the hydroxylation of a series of alkyl aryl ketones (Figure 11; Note: to avoid confusion, the imine isomer in which the directing group is facing the alkylic substituent is defined as the A isomer and the imine isomer in which the directing group is facing the aromatic ring is defined as the B isomer).In one of our previous reports, we described that the Cu-directed hydroxylation of benzaldehyde using 2-picolylamine and H 2 O 2 led to selective γ sp 2 C−H hydroxylation (see entries 1 in Figure 11). 18Conversely, the substrate−ligand derived from 2-picolylamine and cyclohexyl phenyl ketone underwent selective β sp 3 C−H hydroxylation (entry 16, Figure 11).The regioselectivity of these transformations was attributed to the formation of one of the imine isomers (B isomer for benzaldehyde and A isomer for cyclohexyl phenyl ketone, see Figure 11).
For this article, we synthesized substrate−ligands derived from acetophenone, aryl-substituted propiophenones, arylsubstituted 2-phenylacetophenones, isovalerophenone, and 2,2-dimethylpropiophenone (see Figure 11).Variations in the size of the alkyl substituent led to changes in the substrate− ligand imine A/B ratios, in agreement with the findings described by Boyd and co-workers (see Figure 2A).For The Journal of Organic Chemistry example, acetophenone produced the imine B isomer almost selectively ( 4XAlk L 4R-py A / 4XAlk L 4R-py B : 6/94, entry 2 in Figure 11), while propiophenone produced substantial amounts of the imine isomer A ( 4XAlk L 4R-py A / 4XAlk L 4R-py B : 32/68, entry 3 in Figure 11).The oxidation of the acetophenone substrate−ligand led to the selective formation of the sp 2 hydroxylation product (entry 2 in Figure 11), while the oxidation of the propiophenone substrate−ligand formed products derived from the hydroxylation of the β sp 3 C−H and γ sp 2 C−H bonds (entry 3 in Figure 11; Note: see the SI for details on the characterization and quantification of the products derived from β sp 3 C−H hydroxylation, which included diketone and benzoic acid as overoxidation products).Substitution of the para position of the aromatic ring of the propiophenone substrate led to substantial changes in the 4XAlk PL 4R-py A / 4XAlk PL 4R-py B ratios, with the MeO substituent favoring aromatic sp 2 hydroxylation ( 4XAlk PL 4R-py A / 4XAlk PL 4R-py B : 30/70, entry 4 in Figure 11) w

h e n c o m p a r e d w i t h t h e u n s u b s t i t u t e d ( 4XAlk PL 4R-py
A / 4XAlk PL 4R-py B : 37/63, entry 3 in Figure 11) and the 4-F substituted systems ( 4XAlk PL 4-py A / 4XAlk PL 4-py

B
: 47/53, entry 5 in Figure 11).Interestingly, variations on the directing group also led to changes in the 4XAlk PL 4R-py A / 4XAlk PL 4R-py B ratio for the propiophenone substrate, with 2-(aminomethyl)-4m e t h o x y p y r i d i n e e n h a n c i n g s p 2 h y d r o x y l a t i o n ( 4XAlk PL 4R-py A / 4XAlk PL 4R-py B : 29/71, entry 6 in Figure 11) w h e n c o m p a r e d w i t h 2 -p i c o l y l a m i n e ( 4XAlk PL 4R-py A / 4XAlk PL 4R-py B : 37/63, entry 3 in Figure 11) and 2-(aminomethyl)-4-chloropyridine diminishing sp 3 hydroxylation ( 4XAlk PL 4R-py A / 4XAlk PL 4R-py B : 39/61, entry 7 in Figure 11).The Cu-directed hydroxylation of the substrate−ligands derived from 2-phenylacetophenones was also analyzed (entries 8−12 in Figure 11).When compared to the propiophenone systems, we observed a higher tendency for sp 3 hydroxylation (i.e., 4XAlk PL 4R-py A / 4XAlk PL 4R-py B : 51/49 for 2-phenylacetophe-Figure 11.Cu-promoted hydroxylation of systems derived from acyclic alkyl aryl ketones and varying directing groups using H 2 O 2 as an oxidant (see the SI for experimental details on the synthesis of substrate−ligands and hydroxylation reactions).Note: see the SI for details on the characterization and quantification of the products derived from β sp 3 C−H oxidation, including hydroxylation and overoxidation products.
The Journal of Organic Chemistry none vs 4XAlk PL 4R-py A / 4XAlk PL 4R-py B : 37/63 for propiophenone), which was attributed to the weakness of the β sp 3 C−H bond in 2-phenylacetophenone (benzylic C−H bond) when compared to propiophenone (see entries 3 and 8 in Figure 11).While the introduction of a methoxy substituent in the para position of the 2-phenylacetopheone aromatic substrate led only to a modest e n h a n c e m e n t o f t h e s p 2 h y d r o x y l a t i o n ( 4 X A l k P L 4 R -p y A / 4 X A l k P L 4 R -p y B : 4 8 / 5 2 v s 4XAlk PL 4R-py A / 4XAlk PL 4R-py B : 51/49 for unsubstituted, see entries 8 and 9 in Figure 11), the use of an F substituent led to a remarkable enhancement of the sp 3 : 51/49 for unsubstituted, see entries 8 and 10 in Figure 11).Variations on the electronics of the directing group had a small impact on the sp 3  : 51/49 for unsubstituted DG, see entries 8 and 12 in Figure 11).
The hydroxylation of isovalerophenone allowed us to study the competition between the sp 3 and sp 2 C−H bonds (i.e., hydroxylation of isomer A or isomer B) and the competition between β and γ hydroxylation within the sp 3 C−H bonds of isomer A (see entries 13−15 in Figure 11).When compared with propiophenone, we observed that the hydroxylation of the isovalerophenone system led to similar sp 3 /sp 2 ratios ( 4 X A l k P L 4 R -p y A / 4 X A l k P L 4 R -p y B : 2 5 / 7 5 , 4 X A l k P L 4 R -p y A / 4 X A l k P L 4 R -p y B : 2 5 / 7 5 v s 4XAlk PL 4R-py A / 4XAlk PL 4R-py B : 37/63 for propiophenone, see entries 3 and 13 in Figure 11).However, we observed significant differences in the ratio of β and γ hydroxylation within isomer A, with propiophenone being oxidized exclusively at the β position (β/γ: 100/0) and isovalerophenone favoring γ oxidation (β/γ: 7/93).By changing the directing group, we observed modest variations on the sp 3 /sp 2 and β/γ ratios, with 2-(aminomethyl)-4-methoxypyridine enhancing sp 2

KETONES
In one of our previous reports, we showed that the hydroxylation of the substrate ligand derived from 2-picolylamine and 2methyl-1-tetralone occurred selectively at the γ sp 2 C−H bond (entry 1, Figure 12).We were puzzled by this result since we expected the formation of the imine isomer A (DG facing toward the sp 3 C−H bond, see 4XCy L 4R-py A in Figure 12), which would undergo selective β sp 3 hydroxylation like the cyclohexyl phenyl ketone system (see entry 16 in Figure 11).For this paper, we analyzed the 2-methyl-1-tetralone system again, and we observed that the installation of the 2-picolylamine led to the selective formation of the imine isomer B, which explained our prior hydroxylation results (see NMR analysis in the SI).We hypothesized that the rigidity of the cyclic substrate might The Journal of Organic Chemistry preclude the formation of the imine isomer A. To test our hypothesis, we carried out the hydroxylation of a series of cyclic alkyl aryl ketones (Figure 12).In fact, the imine substrate− ligands derived from substituted tetralones (entries 1−4 in Figure 12) and 1-indanone (entry 5 in Figure 12) were all found to produce only the B isomer in solution.Consequently, the hydroxylation of these substrate−ligands selectively produced the oxidation products derived from γ sp 2 C−H functionalization.
Conversely, the substrate−ligands derived from 1-benzosuberone produce two imine isomers (entries 6 to 8 in Figure 12).For all of the DGs tested, we observed similar imine isomer ratios ( 4XCy L 4R-py

■ CONCLUSIONS
To the best of our knowledge, this is the first report in which the regioselectivity of metal-promoted functionalization of C−H bonds using directing groups can be controlled by tuning the electrophilicity of the metal species involved in substrate oxidation.As we have shown, the ability of the imine substrate−ligands derived from unsymmetrical benzophenones to isomerize before the rate-determining step of the reaction allowed for the formation of an electrophilic oxidant (CuOOH core), which can selectively carry out the hydroxylation the electron-rich ring.The electrophilicity of the reactive copper-(II)-hydroperoxo intermediate (which can be benchmarked by measuring the redox potential of the cuprous complexes) can be further modulated by introducing electron-donating (MeO) or electron-withdrawing (Cl, CF 3 ) substituents at the 4-position of the pyridine of the directing group, leading to remarkable regioselectivities (up to 91:9).These findings challenge our initial hypothesis (and the current paradigm in metal-promoted C−H functionalization reactions with imine directing groups), in which the regioselectivity is exclusively determined by the substrate/directing group composition (the E/Z ratio of the imine substrate−ligand) and not by the reactivity of the metalbased intermediates.
Based on the findings on Cu-directed hydroxylation of unsymmetrical benzophenones, we expanded this approach to substrate−ligands derived from unsymmetrical alkyl aryl ketones, and we observed that, like in unsymmetrical benzophenones, the regioselectivity (sp 3 vs sp 2 ) did not only depend on the ratio of the imine substrate−ligands but also on the electronics of the substrate (e.g., inclusion of MeO groups on the aryl substituent enhanced sp 2 hydroxylation) and the electronics of the directing group (e.g., use of electron-poor directing groups enhanced sp 3 hydroxylation).We believe that the results presented in this article will not only provide novel synthetic tools for the selective functionalization of complex molecules using cheap reagents under mild conditions (Cu, H 2 O 2 at room temperature) but also a better understanding of the reaction mechanisms by which Cu-dependent monooxygenase and peroxygenase metalloenzymes (e.g., lytic polysaccharide monooxygenases 16,33−35 ) and mononuclear Cu/O 2 synthetic inorganic complexes 36−38 perform selective C−H oxidations.
■ EXPERIMENTAL SECTION Materials.All reagents and solvents were purchased at the highest level of purity and used as received except as noted.Solvents were purified and dried by passing through an activated alumina purification system (mBRAUN SPS) or by conventional distillation techniques.
Physical Methods.The synthesis of copper complexes and preparation of some NMR samples were carried out under anaerobic conditions in an mBRAUN MB-Unilab Pro SP Glovebox system.All NMR experiments were collected at 300 K on either a two-channel Bruker Avance III NMR instrument equipped with a Broad Band Inverse (BBI) probe or a Bruker NEO 500 NMR spectrometer equipped with a multinuclear BBO Prodigy cryoprobe.Both instruments operate at 500 MHz for 1 H (125.7 MHz for 13 C{ 1 H}).The 1 H NMR spectra are referenced to residual protio solvents (7.26 ppm for CDCl 3 and 1.94 ppm for CD 3 CN), and the 13 C{ 1 H} NMR spectra are referenced to CDCl 3 (77.2ppm).Structural assignments were made with additional information from COSY, NOESY, HMBC, and HSQC experiments.ESI-MS high-resolution mass spectrometry was performed on a Thermo Scientific Exactive Plus EMR Orbitrap Mass Spectrometer in the Department of Chemistry at Carnegie Mellon University.Electrochemical measurements were carried out on a model 620E Electrochemical Workstation (CH Instruments) using a glassy carbon working electrode (4 mm diameter), a Pt wire as the counter electrode, and Ag/AgNO 3 (0.01 M in CH 3 CN) as the reference electrode.The working electrode was polished before each measurement with alumina polishing powder onto a wet polishing cloth.All measurements were made in CH 3 CN with 1 mM Cu complex and NBu 4 PF 6 as the electrolyte (100 mM) at room temperature.
General Procedure for the Synthesis of Imine Substrate− Ligands.In an oven-dried flask, 2-picolylamine (2.2 equiv) was added to the benzophenone substrate (9.85 mmol) and p-toluenesulfonic acid monohydrate (cat.20 mg) in toluene (50 mL).The reaction mixture was refluxed under argon with a Dean−Stark apparatus until imine formation was completed.The reaction was cooled to room temperature and diluted with diethyl ether (30 mL).The organic layer was washed with saturated ammonia chloride (20 mL × 2), saturated aqueous sodium bicarbonate (20 mL), and brine (20 mL) and dried with sodium sulfate.The final product was isolated under vacuum.The purity of the resulting imine substrate−ligands was analyzed by 1 H NMR by adding a known amount of an internal standard (1,3,5trimethoxybenzene or 1,2-dichloroethane).
Standard Procedure for the Hydroxylation of the Imine Substrate−Ligands.In the glovebox, 4 mL of acetone was added to a 20 mL vial containing 0.159 mmol of the imine substrate−ligand equipped with a stir bar.To the solution, 0.159 mmol of [Cu I (CH 3 CN) 4 ](PF 6 ) was added and allowed to react.The solution mixture was taken out of the glovebox, and 5 equiv of 30% H 2 O 2 were added.After 30 min, the reaction was quenched using Na 2 EDTA (50 mL, pH = 4).The resulting mixture was extracted with EtOAc (50 mL × 3).The organic phases were separated, combined, dried over MgSO 4 , filtered, and dried under vacuum.The reaction products were dissolved in 1.4 mL of CDCl 3 solution containing 27.1 mg of 1,3,5trimethoxybenzene (internal standard).The reaction products were quantified by 1 H NMR using integration signals that correspond to the starting material and products with the integration signal of the internal standard.
Standard Procedure for the Removal of the Directing Group.After hydroxylation, the products were dissolved in a round-bottom flask with 50 mL of EtOAc, and 100 mL of 1 M HCl was added.The resulting mixture was stirred at room temperature for 30 min and was extracted with EtOAc (50 mL × 2).The organic phases were separated, The Journal of Organic Chemistry

Figure 3 .
Figure 3. NMR analysis of the Cu-directed hydroxylation of unsymmetrical benzophenones (Note: as an example, we included the data for 4fluorobenzophenone). Characterization and quantification of the reaction products were carried out in CDCl 3 using 2,4,6-trimethoxybenzene as an internal standard (I.S.).See the SI for further details.

Figure 5 .
Figure 5. Mechanistic proposal for the Cu-directed hydroxylation of substrate−ligands derived from unsymmetrical benzophenones.

Figure 6 .
Figure 6.Determination of the E/Z ratio of ligands, cuprous complexes, and hydroxylation products under different reaction conditions.

Figure 7 .
Figure 7. Electrophilicity of the CuOOH cores derived from 4,4′-disubstituted benzophenones18 (A) and 4-substituted benzophenones (B) with 2picolylamine as a directing group.Note: the Hammett plot included in panel B was constructed by plotting the relative reaction rates for the hydroxylation of the substituted (k X ) and unsubstituted (k H ) rings (k X /k H obtained from the 4X PL E / 4X PL Z ratios) vs the σ p values.The electrophilicity plot included in panel B was constructed by plotting k X /k H vs the E 1/2 values of the Cu I complexes derived from 2-picolylamine and 4-substituted benzophenones.

Figure 8 .
Figure 8. (A) Cu-promoted hydroxylation of systems derived from 4,4′-disubstituted benzophenones and 2-picolylamine using H 2 O 2 as an oxidant (see the SI for experimental details on the synthesis of substrate−ligands and hydroxylation reactions).Note: the regioselectivity results ( 4×4 ′ Y PL A / 4×4 ′ Y PL B column) are color-coded according to orange (results obtained for 4-substituted benzophenones), red (worsened regioselectivity for 4,4′-disubstituted benzophenones when compared to 4-substituted benzophenone analogue), or green (improved regioselectivity for 4,4′disubstituted benzophenones when compared to 4-substituted benzophenone analogue).(B) Electrophilicity of the CuOOH cores involved in the hydroxylation of systems derived from unsymmetrical 4,4′-disubstituted benzophenones, 2-picolylamine, and H 2 O 2 .Note: the Hammet plot included in (B) was constructed by plotting the relative reaction rates between the hydroxylation of the arene ring containing the MeO substituent (k MeO ) and the hydroxylation of the other arene ring (k Y , Y: Me, H, F, Cl, CF 3 ); (k MeO /k Y , calculated from the 4MeO4 ′ Y PL A / 4MeO4 ′ Y PL B ratios) vs the difference between the σ p value of MeO and the σ p value of other substituents.The electrophilicity plot was constructed by plotting k MeO /k Y vs the difference in the E 1/2 values of the Cu I complexes derived from 2-picolylamine and 4,4′-dimethoxybenzophenone or unsymmetrical 4,4′-disubstituted benzophenone.
I and H 2 O 2 to produce hydroxylation p r o d u c t s w i t h r e m a r k a b l e r e g i o s e l e c t i v i t y ( 4MeO4 ′ F PL A / 4MeO4 ′ F PL B : 76/24).Similar regioselectivities were observed in the hydroxylation of4 M e O 4 ′ C l L ( 4 M e O 4 ′ C l P L A / 4 M e O 4 ′ C l P L B : 7 5 / 2 5 ) a n d 4 M e 4 ′ F L ( 4Me4 ′ F PL A / 4Me4 ′ F PL B : 74/26).Like we observed in the 4substituted benzophenones, the introduction of the electronwithdrawing CF 3 substituent in 4MeO4 ′ CF3 L enhanced significantly the selectivity for the hydroxylation of the electron-rich arene ring ( 4MeO4 ′ CF3 PL A / 4MeO4 ′ CF3 PL B : 86/14).Conversely, the intramolecular competition between the 4-MeO-substituted and 4-Me-substituted arene rings in 4MeO4 ′ Me L led to a slight e x c e s s o f o n e o f t h e h y d r o x y l a t i o n p r o d u c t s ( 4MeO4 ′ Me PL A / 4MeO4 ′ Me PL B : 53/47).
4MeO4 ′ Cl PL 4MeO-py A / 4MeO4 ′ Cl PL 4MeO-py B : 73/ 2 7 ) o r 2 -( a m i n o m e t h y l ) -4 -c h l o r o p y r i d i n e ( 4MeO4 ′ Cl PL 4Cl-py A / 4MeO4 ′ Cl PL 4Cl-py B : 74/26).A slight decrease in the regioselectivity was found in the hydroxylation of 4methyl-4′-fluorobenzophenone when 2-(aminomethyl)-4-methoxypyridine was used when compared with 2-picolylamine (entries 4 and 5 in Figure 10).Conversely, the use of 2-(aminomethyl)-4-(trifluoro)methylpyridine in the hydroxylation of 4-methoxy-4′-(trifluoro)benzophenone led to the h i g h e s t r e g i o s e l e c t i v i t y r e s u l t s i n t h i s a r t i c l e ( 4MeO4 ′ CF3 PL 4CF3-py A / 4MeO4 ′ CF3 PL 4CF3-py B : 91/9).

Figure 10 .
Figure 10.Cu-promoted hydroxylation of systems derived from unsymmetrical 4,4′-disubstituted benzophenones and varying directing groups using H 2 O 2 as an oxidant (see the SI for experimental details on the synthesis of substrate−ligands and hydroxylation reactions).Note: the regioselectivity results ( 4×4 ′ Y PL 4R-py A / 4×4 ′ Y PL 4R-py Z column) are color-coded according to orange (results obtained with 2-picolylamine and results with other DGs that did not improve the regioselectivity when compared to 2-picolylamine), red (worsened regioselectivity for other DGs when compared to 2picolylamine), or green (improved regioselectivity for other DGs when compared to 2-picolylamine).

Figure 12 .
Figure 12.Cu-promoted hydroxylation of systems derived from cyclic alkyl aryl ketones and 2-picolylamine using H 2 O 2 as an oxidant (see the SI for experimental details on the synthesis of substrate−ligands and hydroxylation reactions).Note: see the SI for details on the characterization and quantification of the products derived from b sp3 C−H hydroxylation.

A/
4XCy L 4R-py B ∼ 35:65), but the ration of sp 3 / sp 2 varied.While 2-picolylamine and 2-(aminomethyl)-4m e t h o x y p y r i d i n e f a v o r e d s p 2 h y d r o x y l a t i o n ( 4XCy PL 4R-py A / 4XCy PL 4R-py B : 25/75 and 24/76, respectively), the use of 2-(aminomethyl)-4-chloropyridine allowed for r e a c h i n g h i g h e r s p 3 h y d r o x y l a t i o n y i e l d s ( 4XCy PL 4R-py A / 4XCy PL 4R-py B : 46/54).These findings are aligned with the selectivity results obtained for all of the acyclic alkyl aryl ketones analyzed (propiophenones, aryl-substituted 2-phenylacetophenones, and isovalerophenone), in which utilization of 2-(aminomethyl)-4-methoxypyridine enhanced sp 2 hydroxylation, while 2-(aminomethyl)-4-methoxypyridine enhanced sp 2 hydroxylation, suggesting that electron-rich CuOOH cores are prone to oxidize sp 2 C−H bonds and electron-poor ones prefer sp 3 C−H bonds.