Electronic and Torsional Effects on Hydrogen Atom Transfer from

Nov 15, 2017 - A kinetic study on the hydrogen atom transfer (HAT) reactions from the aliphatic C–H bonds of a series of 1-Z-pentyl, 1-Z-propyl, and...
4 downloads 16 Views 1003KB Size
Download PDF
Article Cite This: J. Org. Chem. 2017, 82, 13542−13549

pubs.acs.org/joc

Electronic and Torsional Effects on Hydrogen Atom Transfer from Aliphatic C−H Bonds: A Kinetic Evaluation via Reaction with the Cumyloxyl Radical Michela Salamone,† Teo Martin,† Michela Milan,‡ Miquel Costas,*,‡ and Massimo Bietti*,† †

Dipartimento di Scienze e Tecnologie Chimiche, Università “Tor Vergata”, Via della Ricerca Scientifica, 1 I-00133 Rome, Italy QBIS Research Group, Institut de Química Computacional i Catàlisi (IQCC) and Departament de Química, Universitat de Girona, Campus Montilivi, Girona E-17071, Catalonia, Spain

‡

S Supporting Information *

ABSTRACT: A kinetic study on the hydrogen atom transfer (HAT) reactions from the aliphatic C−H bonds of a series of 1-Z-pentyl, 1-Zpropyl, and Z-cyclohexyl derivatives and of a series of N-alkylamides and N-alkylphthalimides to the electrophilic cumyloxyl radical (CumO•) has been carried out. With 1-pentyl and 1-propyl derivatives, α-CH2 activation toward CumO• is observed for Z = Ph, OH, NH2, and NHAc, as evidenced by an increase in kH as compared to the unsubstituted alkane substrate. A decrease in kH has been instead measured for Z = OAc, NPhth, CO2Me, Cl, Br, and CN, indicative of α-CH2 deactivation with HAT that predominantly occurs from the most remote methylenic site. With cyclohexyl derivatives, α-CH activation is only observed for Z = OH and NH2, indicative of torsional effects as an important contributor in governing the functionalization selectivity of monosubstituted cyclohexanes. In the reactions of N-alkylamides and N-alkylphthalimides with CumO•, the reactivity and selectivity patterns parallel those observed in the oxidation of the same substrates with H2O2 catalyzed by manganese complexes, supporting the hypothesis that both reactions proceed through a common HAT mechanism. The implications of these findings and the potential of electronic, stereoelectronic, and torsional effects as tools to implement selectivity in C−H oxidation reactions are briefly discussed.

■

INTRODUCTION Selective functionalization of aliphatic C−H bonds represents an important goal of modern synthetic organic chemistry, because these reactions can offer advantages both in terms of decreased waste generation and reaction step economy.1−4 Within this framework, differentiating between the C−H bonds of organic molecules with high levels of regio- and stereoselectivity and between the C−H bonds of the substrate and of the functionalized product remains, however, an ongoing challenge, and accordingly, the development of procedures for the accomplishment of these rewarding synthetic goals represents an area of intense investigation. Among the methodologies that have been developed for this purpose, procedures based on hydrogen atom transfer (HAT) from aliphatic C−H bonds to radical and radical-like reagents such as metal−oxo species have attracted considerable attention. The factors that govern reactivity and site-selectivity have been discussed in detail and include bond strengths; electronic (also named inductive or polar), steric, and stereoelectronic effects; conjugation and hyperconjugation; and, with cyclohexane derivatives, torsional effects.5,6 Medium effects have also emerged as a powerful tool that has been successfully employed to dramatically alter both reactivity and © 2017 American Chemical Society

site selectivity in HAT-based C−H functionalization procedures.7−12 On the basis of the electrophilic nature of most of the commonly employed hydrogen-abstracting reagents, HAT will preferentially occur from an electron-rich and thus activated C−H bond rather than from an electron-poor and deactivated one of similar strength, pointing toward electronic effects as an extremely powerful tool to govern site selectivity in these processes.5,7b,13 Accordingly, linear aliphatic substrates bearing a terminal functional group Z and aliphatic substrates bearing two tertiary C−H bonds in proximal and remote positions with respect to a functional group Z (Scheme 1) are customarily employed as probes for studying the role played by electronic effects on the C−H functionalization selectivity.14−24 Despite the widespread application of such probes to the study of electronic effects, a thorough evaluation of the nature of the functional group on the activation or deactivation of proximal and remote aliphatic C−H bonds toward HAT is still lacking. For this purpose, in order to obtain quantitative information on the effect of a functional group Z on the HAT Received: October 19, 2017 Published: November 15, 2017 13542

DOI: 10.1021/acs.joc.7b02654 J. Org. Chem. 2017, 82, 13542−13549

Article

The Journal of Organic Chemistry Scheme 1

reactivity of aliphatic C−H bonds, we felt it of interest to carry out a detailed time-resolved kinetic study on the reactions of a prototypical electrophilic HAT reagent such as the cumyloxyl radical [PhC(CH3)2O•, CumO•] with an extended series of 1pentyl, 1-propyl, and cyclohexyl derivatives, differing by the nature of the substituent Z, the structures of which are displayed in Chart 1. Chart 1 Figure 1. Plots of the observed rate constant (kobs) against [CH3CH2CH2CH2CH2Z] for the reactions of the cumyloxyl radical (CumO•) measured in argon-saturated MeCN solution at T = 25 °C by following the decay of CumO• at 490 nm. Linear regression analysis: hexanenitrile (Z = CN, black circles), intercept = 7.29 × 105 s−1, kH = 1.77 × 105 M−1 s−1, r2 = 0.9967; methyl hexanoate (Z = CO2Me, white circles), intercept = 7.68 × 105 s−1, kH = 1.91 × 105 M−1 s−1, r2 = 0.9944; pentane (Z = H, gray circles), intercept = 7.71 × 105 s−1, kH = 2.97 × 105 M−1 s−1, r2 = 0.9962; 1-phenylpentane (Z = Ph, black squares), intercept = 7.15 × 105 s−1, kH = 8.10 × 105 M−1 s−1, r2 = 0.9977; N-pentylacetamide (Z = NHAc, white squares), intercept = 8.02 × 105 s−1, kH = 9.63 × 105 M−1 s−1, r2 = 0.9996; 1-pentanol (Z = OH, gray squares), intercept = 7.02 × 105 s−1, kH = 1.52 × 106 M−1 s−1, r2 = 0.9993; pentylamine (Z = NH2, black diamonds), intercept = 7.81 × 105 s−1, kH = 1.56 × 107 M−1 s−1, r2 = 0.9982.

In keeping with the results of our own recent study on the oxidation reactions of a series of N-alkylamides and Nalkylphthalimides (S1−S12) with hydrogen peroxide catalyzed by manganese complexes,15 and in order to draw a comparison between the reactivity of a high-valent manganese−oxo species and an oxygen-centered radical toward aliphatic C−H bonds, the reactions with CumO• have been also extended to this group of substrates, the structures for which are displayed below in Table 4 and Schemes 4 and 5. For the sake of clarity, substrate numbering used in our previous study has been maintained in the present work.

■

derivatives bearing electron-withdrawing groups (EWGs), the plots for Z = OAc, NPhth, Cl, and Br have not been included. Additional kobs vs [substrate] plots for the reactions of CumO• with the other 1-pentyl derivatives, for the 1-propyl and cyclohexyl derivatives, and for N-alkylamides and N-alkylphthalimides S1−S12 are displayed in the Supporting Information as Figures SI1−SI7. The kH values thus obtained for reaction of CumO• with the 1-pentyl, 1-propyl, and cyclohexyl derivatives displayed in Chart 1 are collected in Tables 1, 2, and 3, respectively. Those obtained for reaction of CumO• with amides and phthalimides S1−S12 are collected in Table 4. HAT from pentane, 1-pentanol, and pentylamine to CumO• has been previously investigated and discussed.7a By considering that unactivated methyl groups display an extremely low reactivity toward tert-alkoxyl radicals (kH ≤ 1.3 × 104 M−1 s−1 per methyl group),26 the kH value measured for reaction of CumO• with pentane (kH = 3.1 × 105 M−1 s−1) indicates that with this substrate HAT mostly occurs from the methylene groups. On the basis of this observation, the ∼5- and ∼50-fold increase in kH measured on going from pentane to 1-pentanol and pentylamine, respectively, is indicative of α-CH2 activation determined by the presence of the electron-rich OH and NH2 groups, with HAT that for both substrates predominantly occurs from this site. α-C−H activation toward HAT to CumO• reflects the contribution of both kinetic (electronic effects) and thermodynamic components (bond weakening determined by hyperconjugative overlap between the C−H σ* orbital and a heteroatom lone pair), with the one-order of magnitude increase in kH observed on going from the alcohol to

RESULTS AND DISCUSSION CumO• was generated by 355 nm laser flash photolysis (LFP) of argon-saturated acetonitrile solutions (T = 25 °C) containing 1.0 M dicumyl peroxide. CumO• displays a broad absorption band in the visible region of the spectrum that in this solvent is centered at 485 nm, and in the absence of hydrogen atom donor substrates decays almost exclusively by C−CH3 βscission.25 This spectroscopic feature makes the direct measurement of HAT rate constants by nanosecond LFP particularly convenient. Product studies on the reactions of CumO• with hydrogen atom donor substrates are, on the other hand, very limited, because of complications related to trapping of the carbon-centered radical formed by HAT and follow-up transformation of the first formed C−H functionalization products.7a Time-resolved kinetic studies on the reactions of CumO• with the different substrates shown in Chart 1 and with amides and phthalimides S1−S12 were carried out by employing the LFP technique. When the observed rate constants (kobs), measured following the decay of the CumO• visible absorption band, were plotted against substrate concentration, excellent linear relationships were observed. The second-order rate constants for HAT to CumO• (kH) were obtained from the slope of these plots. As an example, Figure 1 shows the kobs vs [substrate] plots for the reactions of 1-pentyl derivatives CH3CH2CH2CH2CH2Z (Z = H, CN, CO2Me, Ph, NHAc, OH, NH2) with CumO•, for measurements carried out in MeCN at T = 25 °C. For the sake of clarity and because of the very similar kH values measured for the reactions of 1-pentyl 13543

DOI: 10.1021/acs.joc.7b02654 J. Org. Chem. 2017, 82, 13542−13549

Article

The Journal of Organic Chemistry Table 1. Second-Order Rate Constants for Reaction of the Cumyloxyl Radical (CumO•) with 1-Pentyl Derivatives (CH3CH2CH2CH2CH2Z), Measured in Argon-Saturated Acetonitrile Solution at T = 25 °Ca kH/M−1 s−1

substrate

Z

pentane 1-phenylpentane 1-pentanol 1-pentyl acetate pentylamine N-pentylacetamidec N-pentylphthalimided methyl hexanoate 1-chloropentane 1-bromopentane hexanenitrile

H Ph OH OAc NH2 NHAc NPhth CO2Me Cl Br CN

(3.1 ± 0.1) (7.9 ± 0.2) (1.46 ± 0.07) (2.06 ± 0.08) (1.55 ± 0.02) (9.58 ± 0.06) (2.3 ± 0.2) (1.97 ± 0.06) (1.84 ± 0.08) (1.76 ± 0.03) (1.82 ± 0.04)

× × × × × × × × × × ×

kH(subst)/kH(pentane) 105 b 105 106 b 105 107 b 105 105 105 105 105 105

1.0 2.54 4.7 0.66 50 3.1 0.74 0.64 0.59 0.57 0.59

Measured in argon-saturated acetonitrile solution at T = 25 °C employing 355 nm LFP: [dicumyl peroxide] = 1.0 M. Values of kH were determined from the slope of the kobs vs [substrate] plots, where in turn kobs values were measured following the decay of the CumO• visible absorption band at 490 nm. Average of at least two determinations. bReference 7a. cNumbered as S2 in Table 4; see below. dNumbered as S4 in Table 4; see below. a

Table 2. Second-Order Rate Constants for Reaction of the Cumyloxyl Radical (CumO•) with 1-Propyl Derivatives (CH3CH2CH2Z), Measured in Argon-Saturated Acetonitrile Solution at T = 25 °Ca substrate 1-phenylpropane 1-propanol 1-propyl acetate propylamine N-propylacetamide N-propylphthalimide methyl butanoate 1-chloropropane 1-bromopropane butanenitrile

X

kH/M−1 s−1

Ph OH OAc NH2 NHAc NPhth CO2Me Cl Br CN

(6.16 ± 0.02) × 10 (1.04 ± 0.04) × 106 ≤3 × 104 c (1.04 ± 0.06) × 107 d (5.7 ± 0.3) × 105 ≤3 × 104 c ≤3 × 104 c (4.2 ± 0.2) × 104 (4.0 ± 0.1) × 104 ≤3 × 104 c

kH(subst)/kH(propane)b 5

6.2 10.4 ≤0.3 104 5.7 ≤0.3 ≤0.3 0.42 0.40 ≤0.3

a Measured in argon-saturated acetonitrile solution at T = 25 °C employing 355 nm LFP: [dicumyl peroxide] = 1.0 M. Values of kH were determined from the slope of the kobs vs [substrate] plots, where in turn kobs values were measured following the decay of the CumO• visible absorption band at 490 nm. Average of at least two determinations. bBy assuming that kH(propane) ∼ 1.0 × 105 M−1 s−1. cOnly an upper limit to kH could be determined. dReference 32.

Table 3. Second-Order Rate Constants for Reaction of the Cumyloxyl Radical (CumO•) with Cyclohexyl Derivatives, Measured in Argon-Saturated Acetonitrile Solution at T = 25 °Ca substrate

Z

cyclohexane phenylcyclohexane cyclohexanol cyclohexyl acetate cyclohexylamine N-cyclohexylacetamide N-cyclohexylphthalimide methyl cyclohexanecarboxylate chlorocyclohexane bromocyclohexane cyclohexanecarbonitrile

H Ph OH OAc NH2 NHAc NPhth CO2Me Cl Br CN

kH/M−1 s−1 (1.1 ± 0.1) (9.1 ± 0.2) (2.66 ± 0.05) (4.2 ± 0.2) (2.1 ± 0.1) (6.9 ± 0.1)

kH(subst)/kH(cyclohex)

× × × × × ×

106b 105 106 c 105 107 c 105

1.0 0.83 2.42 0.38 19.1 0.63

× × × ×

105 105 105 105

0.56 0.42 0.30 0.35

d

(6.14 ± 0.02) (4.62 ± 0.04) (3.3 ± 0.1) (3.9 ± 0.1)

Measured in argon-saturated acetonitrile solution at T = 25 °C employing 355 nm LFP: [dicumyl peroxide] = 1.0 M. kH values were determined from the slope of the kobs vs [substrate] plots, where in turn kobs values were measured following the decay of the CumO• visible absorption band at 490 nm. Average of at least two determinations. bReference 28. cReference 7a. dSolubility was too low to allow the determination of the kH value. a

contribution to kH of HAT from a single nonactivated CH2 group to CumO• of ∼1 × 105 M−1 s−1, derived from the kH value measured for reaction of CumO• with pentane.29 Among the substrates displayed in Table 1, α-C−H activation is also observed for 1-phenylpentane and Npentylacetamide, as clearly shown by the 2.5- and 3.1-fold increase in the measured kH values as compared to that of

the amine that can be accounted for on the basis of the greater inductive electron-withdrawing character of the OH group as compared to NH2.27 This comparison indicates, moreover, that whereas pentylamine undergoes HAT almost exclusively from the α-C−H bonds, with 1-pentanol competitive HAT from the more remote methylenic groups of the aliphatic chain also occurs, an observation that is supported by the estimated 13544

DOI: 10.1021/acs.joc.7b02654 J. Org. Chem. 2017, 82, 13542−13549

Article

The Journal of Organic Chemistry Table 4. Second-Order Rate Constants (kH) for Reaction of the Cumyloxyl Radical (CumO•) with N-Alkylamides and NAlkylphthalimidesa

Scheme 2

former substrate to 1-pentanol and pentylamine, respectively, despite of the significantly weaker α-C−H bonds displayed by 1-phenylpentane (BDE ∼ 86 kcal mol−1 for the benzylic C−H bonds of linear 1-phenylalkanes; BDE = 93.5 and 90.5 kcal mol−1 for the α-C−H bonds of 1-pentanol and pentylamine, respectively).31 As compared to pentane, a decrease in kH has been measured in all cases when Z is an EWG, a behavior that is indicative of electronically induced α-CH2 deactivation toward HAT to the electrophilic CumO•. Interestingly, comparable kH values have been measured for this group of substrates, irrespective of the nature of the Z group (kH between 1.76 and 2.3 × 105 M−1 s−1 for Z = Br, CN, Cl, CO2Me, OAc, NPhth), a behavior that suggests common selectivity patterns and points in all cases toward predominant HAT from the most remote and least electronically deactivated δ-methylenic site. This hypothesis is in full agreement with the results of a number of studies on the reactions of 1-pentyl derivatives bearing an electron-withdrawing substituent with electrophilic radical and radical-like HAT reagents, which have shown in all cases predominant or exclusive formation of the product deriving from δ-C−H bond functionalization.14−16,18,22 In order to minimize the kinetic contribution from HAT from remote sites and thus to better quantify the activation or deactivation of proximal C−H bonds determined by electronic effects, the reactions of CumO• with the corresponding 1propyl derivatives have been studied, the results for which are displayed in Table 2. The reactivity patterns observed for the 1propyl derivatives closely parallel those discussed above for the 1-pentyl ones, with the main difference being represented by the significantly larger decrease in the measured kH values on going from substrates bearing activating substituents (Z = Ph, NHAc, OH, NH2) to those characterized by the presence of deactivating ones (Z = OAc, CO2Me, NPhth, Cl, Br, CN). With most of the latter substrates, only an upper limit to kH could be determined (kH ≤ 3 × 104 M−1 s−1). In keeping with the discussion outlined above, the former substrates undergo HAT almost exclusively or predominantly from the activated α-C−H bonds. The decrease in kH measured for Z = Ph, NHAc, and OH on going from 1-pentyl to 1-propyl derivatives strongly supports the hypothesis outlined above of competitive HAT from the remote methylenic groups of the former substrates. On the other hand, with the substrates for which Z = EWG, the lack of remote and less-deactivated γ- and δ-methylenic groups significantly decreases the HAT reactivity toward CumO•, which now will eventually occur from the leastdeactivated β-C−H bonds. Taken together, these observations provide full support to the hypothesis that with 1-pentyl derivatives bearing EWG substituents HAT predominantly occurs from the most remote and least electronically deactivated δ-methylenic sites and to a lesser extent γmethylenic sites. The kH values measured for HAT from the cyclohexyl derivatives to CumO• are displayed in Table 3.

Measured in argon-saturated acetonitrile solution at T = 25 °C employing 355 nm LFP: [dicumyl peroxide] = 1.0 M. Values of kH were determined from the slope of the kobs vs [substrate] plots, where in turn kobs values were measured following the decay of the CumO• visible absorption band at 490 nm. Average of at least two determinations. a

pentane. Also with these substrates the relatively low kH(subst)/kH(pentane) ratios indicate that competitive HAT from the more remote methylenic groups occurs. The higher rate constant measured for N-pentylacetamide as compared to that of pentane, together with the 16-fold decrease in k H measured on going from pentylamine to Npentylacetamide, confirms that, despite the decrease in hyperconjugative donation from the nitrogen lone pair to the α-C−H σ* orbital, resulting from the electron-withdrawing character of the acetyl group, the amide nitrogen atom is still sufficiently electron rich to activate the α-C−H bonds toward HAT to the electrophilic CumO•. Comparison between the kH values measured for 1phenylpentane, 1-pentanol, and pentylamine provides a striking example of the important role played by electronic effects in these reactions (Scheme 2). This is clearly evidenced by the ∼2- and 20-fold increase in kH measured on going from the 13545

DOI: 10.1021/acs.joc.7b02654 J. Org. Chem. 2017, 82, 13542−13549

Article

The Journal of Organic Chemistry As compared to cyclohexane, an increase in kH has been only observed in the reactions of CumO• with cyclohexanol and cyclohexylamine [kH(subst)/kH(cyclohexane) = 2.4 and 19.1, respectively], in line with α-C−H activation determined by the presence of the electron-rich OH and NH2 substituents. On the other hand, a slight decrease in kH has been measured for the corresponding reactions of phenylcyclohexane and N-cyclohexylacetamide, for which kH(subst)/kH(cyclohexane) = 0.8 and 0.6, respectively. This behavior is in contrast to that observed with the 1-pentyl and 1-propyl derivatives, where, as compared to the unsubstituted substrate, an increase in kH has been observed for Z = Ph, OH, NH2, and NHAc. A decrease in kH has been measured in all cases on going from cyclohexane to cyclohexyl derivatives characterized by the presence of an EWG substituent (Z = OAc, CO2Me, Cl, Br, CN), with kH(subst)/kH(cyclohexane) ratios between 0.30 and 0.56. Previous studies on the HAT-based functionalization of monosubstituted cyclohexanes have clearly shown that while methylcyclohexane undergoes HAT from all ring positions,23,33 the presence of more bulky substituents, such as t-Bu, Ph, and SiMe3, and of EWG substituents, such as OCOR, NHCOR, CO2H, CO2Me, COMe, and CF3, directs C−H functionalization selectively to positions C3 and C4 of the ring,14b,23,33−36 with a site selectivity, quantified on the basis of the normalized C3/C4 product ratio, that appears to be influenced by the electronic character of the substituent. It has been suggested that bulky electron-releasing (ERG) substituents favor C3 oxidation over C4 (C3/C4 > 1), while prevalent C4 oxidation (C3/C4 < 1) is observed in the presence of EWG substituents.33 However, a recent study on the oxidation of monosubstituted cyclohexanes with hydrogen peroxide catalyzed by manganese complexes has indicated that electronic effects are not the only factor that contributes to defining the functionalization selectivity.34 The absence of products deriving from functionalization at C1 and C2 has been explained on the basis of the contribution of steric and torsional effects. The Z group places the α-C−H in an axial position, and tertiary axial C−H bond deactivation has been recently rationalized on the basis of an increase in torsional strain in the HAT transition state, where planarization of the incipient carbon-centered radical forces the Z group toward an unfavorable eclipsed interaction with the equatorial C−H bonds on the adjacent C2 and C6 positions (Scheme 3,

EWG substituents are also expected to electronically deactivate the C−H bonds at C1 and C2 toward HAT to electrophilic hydrogen-abstracting species. Along these lines, and in keeping with the results of previous studies on the HAT-based functionalization of monosubstituted cyclohexanes mentioned above,14b,23,33−36 the kH values displayed in Table 3 can be rationalized on the basis of the interplay between electronic and torsional effects. The increase in kH measured on going from cyclohexane to cyclohexanol and cyclohexylamine clearly indicates that the ERG character of the OH and NH2 groups overrides tertiary axial C−H bond deactivation, with HAT to CumO• occurring predominantly or almost exclusively from the α-C−H bond (C1). The decrease in kH measured on going from cyclohexane to cyclohexyl derivatives bearing EWG substituents indicates, on the other hand, that the C−H bonds at C1 and C2 (and C6) are deactivated by a combination of electronic and torsional effects, with HAT to CumO• now occurring from the remote and least electronically deactivated C3 and C4 positions. With phenylcyclopentane and N-cyclohexylacetamide, the decrease in kH indicates that torsional effects dominate over electronic effects, with HAT now selectively occurring from the C3 and C4 ring positions. The deactivation at C1 and the associated siteselectivity observed for phenylcyclohexane23,36 is quite striking if one considers in particular that the α-C−H bond is a relatively weak tertiary benzylic bond (BDE = 85.2 kcal mol−1), about 14 kcal mol−1 weaker than the secondary C−H bonds of cyclohexane,31 an observation that points toward tertiary axial C−H bond deactivation as an extremely important contributor in governing the functionalization selectivity of monosubstituted cyclohexanes. The kH values measured for HAT from N-alkylamides and Nalkylphthalimides S1−S12 to CumO• are collected in Table 4. As a matter of comparison, the product distributions that have been recently observed in the C−H bond oxidation of the same substrates with hydrogen peroxide catalyzed by [Mn(CF3SO3)2(dMMpdp)] [dMMpdp = N,N′-bis(2-(3,5-dimethyl-4methoxypyridyl)methyl)-2,2′-bipyrrolidine], Mn−(dMMpdp), a reaction that has been proposed to proceed via HAT to a highvalent manganese−oxo species,15 are displayed in Scheme 4 (for the reactions of the N-alkylpivalamides, acetamides, and trifluoroacetamides) and Scheme 5 (for the reactions of the Nalkylphthalimides).38 Exclusive α-C−H hydroxylation was observed for Npentylamides S1−S3, in line with the activation of this position toward HAT promoted by the amido group. Exclusive hydroxylation at the remote tertiary δ-C−H was observed for N-2-(5-methylhexyl)amides S5−S7, a behavior that was explained on the basis of α-C−H bond deactivation determined by the operation of stereoelectronic effects (see below). Competition between α-C−H and tertiary γ-C−H hydroxylation was observed for N-1-(3-methylbutyl)amides S9−S11. With the latter substrates, α-C−H activation promoted by the pivalamido and acetamido groups was shown to outcompete tertiary γ-C−H hydroxylation, while an opposite behavior was observed for the trifluoroacetamide derivative, in line with the stronger EWG character of this group. With all three N-alkylphthalimides (S4, S8, and S12) exclusive oxidation at the most remote secondary or tertiary C−H bond was observed, in line with the strong EWG character of the phthalimido group. Previous time-resolved kinetic studies on the reactions of CumO• with N-alkylacetamides and N-alkylpivalamides have

Scheme 3

structure A, where X represents a generic hydrogen-atomabstracting radical).37 Deactivation at C2 can be explained accordingly on the basis of the planarization of the incipient carbon radical in the HAT transition state that forces the remaining C−H bond toward an unfavorable eclipsed interaction with the adjacent Z group (Scheme 3, structure B).37 13546

DOI: 10.1021/acs.joc.7b02654 J. Org. Chem. 2017, 82, 13542−13549

Article

The Journal of Organic Chemistry Scheme 4

Among the N-pentyl series, the lowest kH value for HAT to CumO• has been measured for phthalimide S4 (kH = 2.3 × 105 M−1 s−1), where, as discussed above, the strong EWG character of the phthalimido group directs HAT toward the most remote and least electronically deactivated δ-methylenic site. The result of the kinetic studies parallels the exclusive formation of the δketoamide product in the Mn−(dMMpdp)-catalyzed oxidation of S4 (Scheme 5). The increase in kH measured on going from the Npentylamides (S1−S3) to the N-1-(3-methylbutyl) ones (S9− S11) is indicative of an increased contribution to the overall reactivity of HAT from the remote tertiary γ-C−H bond, an observation that is well-supported by the almost identical increase in kH measured within the three substrate couples. The relative importance of the latter contribution, however, is significantly greater for N-(3-methylbutyl)trifluoroacetamide (S11) as compared to the corresponding pivalamide and acetamide substrates S9 and S10. This observation is again in line with the site-selectivity observed in the Mn−(dMMpdp)catalyzed oxidation of these substrates (Scheme 4), where the relative importance of remote γ-C−H functionalization increases in the order S10 < S9 < S11. As described in a recent time-resolved kinetic study on HAT from N-alkylamides to CumO•, a ca. 9-fold decrease in the kH value for HAT from the α-C−H bond was measured on going from N-ethylpivalamide to N-isopropylpivalamide (for which kH = 5.4 × 105 and 5.8 × 104 M−1 s−1, respectively).26b It was proposed that the introduction of a methyl group on the αcarbon increases the energy barrier required to reach the most suitable conformation for HAT, where the α-C−H bond is aligned with the amide π-system, leading to the observed decrease in reactivity. Along these lines, the up to 3-fold decrease in kH measured on going from N-pentylacetamide and pivalamide (S1 and S2) to the corresponding N-(2-(5methylhexyl) ones (S5 and S6) can be explained accordingly on the basis of the deactivation of the tertiary α-C−H bond of the latter substrates toward HAT determined by the operation of stereoelectronic effects. However, as compared to the Mn− (dMMpdp)-catalyzed reactions of S5 and S6, where selective tertiary δ-C−H bond oxidation was observed (Scheme 4), the higher kH values measured for reaction of CumO• with S5 and S6 (kH = 3.4 × 105 and 4.1 × 105 M−1 s−1, respectively) as compared to those measured previously for the corresponding

Scheme 5

shown that HAT from the acetyl methyl group occurs with kH ≤ 1.8 × 104 M−1 s−1, whereas HAT from the pivaloyl tert-butyl group occurs with kH < 1 × 104 M−1 s−1.26b Along this line, the magnitude of the kH values displayed in Table 4 (kH between 2.3 × 105 and 1.08 × 106 M−1 s−1) clearly indicates that with all the amides and phthalimides investigated in the present study HAT from the C−H bonds of the N-alkyl group represents by far the most important reactive pathway. We start the analysis of the kinetic data displayed in Table 4 from the N-pentyl derivatives. Comparison of the kH values measured for S1 and S2 (kH = 7.6 × 105 and 9.58 × 105 M−1 s−1, respectively) with those measured previously for the corresponding reactions of CumO• with N-ethylpivalamide and N-ethylacetamide (kH = 5.4 × 105 and 6.2 × 105 M−1 s−1, respectively),26b both undergoing predominant HAT from the α-C−H bonds of the N-ethyl group, indicates that also with S1 and S2 HAT to CumO• predominantly occurs from the α-C− H bonds. This observation is in line with the site-selectivity observed in the Mn−(dMMpdp)-catalyzed oxidation of these substrates, where the exclusive formation of products deriving from α-C−H hydroxylation was detected (Scheme 4).15 The increase in kH observed on going from the N-ethyl to the Npentyl amides reasonably reflects, however, competitive HAT from the additional methylenic groups of the latter substrates, a behavior that may be indicative of differences in the electrophilic character of CumO• and of the Mn−(dMMpdp)based active oxidant (see later text). 13547

DOI: 10.1021/acs.joc.7b02654 J. Org. Chem. 2017, 82, 13542−13549

The Journal of Organic Chemistry reactions of N-isopropylpivalamide and N-isopropylacetamide (kH = 5.8 × 104 and 1.6 × 105 M−1 s−1, respectively),26b both undergoing predominant HAT from the tertiary α-C−H bond, indicate that with the former substrates HAT from the α-C−H bond to CumO• also contributes to the measured kH value. This different behavior can be reasonably accounted for on the basis of the significantly greater steric hindrance of the Mn− (dMMpdp)-based active oxidant as compared to CumO•. Finally, the kH values measured for reaction of CumO• with the N-alkyltrifluoroacetamides (S3, S7, and S11) and Nalkylphthalimides (S4, S8, and S12) deserve a special comment. The significantly lower value measured for N-pentyltrifluoroacetamide (S3) as compared to the corresponding pivalamide (S1) and acetamide (S2) derivatives (kH = 2.95 × 105, 7.6 × 105, and 9.58 × 105 M−1 s−1, respectively), together with the increase in reactivity observed on going from S3 to the N-(2-(5methylhexyl) and N-1-(3-methyl)butyltrifluoroacetamides (S7 and S11, for which kH = 3.9 × 105 and 4.1 × 105 M−1 s−1, respectively), clearly indicates that with N-alkyltrifluoroacetamides S3, S7, and S11, HAT to CumO• mostly occurs from the remote N-alkyl secondary and tertiary C−H bonds and, only to a minor extent, from the α-C−H bonds. This behavior, although in line with the Mn−(dMMpdp)-catalyzed oxidation of S7 and S11, contrasts with the exclusive formation of the α-C− H hydroxylation product in the oxidation of S3 (Scheme 4). The different behavior observed for CumO• and the Mn− (dMMpdp)-based active oxidant in the reactions of S3 may reflect differences in the electrophilic character of these two HAT reagents, with the former one being more sensitive to electronic α-C−H bond deactivation. Along the same line, the decrease in kH measured on going from S3, S7, and S11 to the corresponding phthalimides (S4, S8, and S12, for which kH = 2.3 × 105, 2.6 × 105, and 2.5 × 105 M−1 s−1, respectively) is in full agreement with the stronger EWG character of the phthalimido group as compared to the trifluoroacetamido one, which increases the extent of α-C−H deactivation, directing HAT toward more remote and less deactivated secondary and tertiary C−H bonds, in full agreement with the site selectivity observed in the oxidation of S4, S8, and S12 catalyzed by Mn−(dMMpdp) (Scheme 5). In conclusion, the reactivity and selectivity patterns observed in the reactions of N-alkylamides and N-alkylphthalimides with the genuine HAT reagent CumO• appear to parallel, at least qualitatively, those observed recently in the oxidation of the same substrates with hydrogen peroxide catalyzed by Mn− (dMMpdp), pointing toward the important role played by electronic and stereoelectronic effects in governing siteselectivity and strongly supporting the hypothesis that the latter reactions proceed through an initial HAT step to a highvalent manganese−oxo species.39 Along the same lines, the reactivity and selectivity patterns observed in the reactions of CumO• with the 1-pentyl, 1-propyl, and cyclohexyl derivatives clearly indicate that electronic and torsional effects can often override C−H bond strengths, suggesting that analogous effects should also operate in HAT reactions from aliphatic C−H bonds of related substrates promoted by manganese−oxo species, and presumably also iron−oxo species, reactions that are generally believed to be governed by C−H bond strengths.41−43 A consequence of the present findings is that the full potential of electronic, stereoelectronic, and torsional effects should be considered to implement selectivity in C−H oxidation reactions of alkyl and cycloalkyl derivatives promoted by iron- and manganese-based catalysts.

Article

■

EXPERIMENTAL SECTION

■

ASSOCIATED CONTENT

Materials. Spectroscopic grade acetonitrile was used in the kinetic experiments. The following substrates were of the highest commercial quality available and were used as received: 1-phenylpentane, 1-pentyl acetate, methyl hexanoate, 1-chloropentane, 1-bromopentane, hexanenitrile, 1-phenylpropane, 1-propanol, 1-propylacetate, methyl butanoate, 1-chloropropane, 1-bromopropane, butanenitrile, phenylcyclohexane, cyclohexanol, cyclohexyl acetate, methyl cyclohexanecarboxylate, chlorocyclohexane, bromocyclohexane, and cyclohexanecarbonitrile. N-Alkylamides S1−S3, S5−S7, S9−S11; N-alkylphthalimides S4, S8, and S12; and N-cyclohexylacetamide were available from previous studies.15,34 N-Propylacetamide was prepared by reaction of propylamine with acetyl chloride according to a previously described procedure, purified by column chromatography (silica gel, eluent hexane−ethyl acetate 4:1), and identified by 1H NMR.44 N-Propylphthalimide was prepared by reaction of 1-bromopropane with the potassium salt of phthalimide according to a previously described procedure,18c purified by column chromatography (silica gel, eluent hexane−ethyl acetate 5:1), and identified by 1H NMR.45 N-Cyclohexylphthalimide was prepared by reaction of cyclohexylamine with phthalic anhydride according to a previously described procedure and identified by 1H NMR.18c Dicumyl peroxide was of the highest commercial quality available and was used as received. Laser Flash Photolysis Studies. LFP experiments were carried out with a laser kinetic spectrometer using the third harmonic (355 nm) of a Q-switched Nd:YAG laser, delivering 8 ns pulses. The laser energy was adjusted to ≤10 mJ/pulse by the use of the appropriate filter. A 3.5 mL Suprasil quartz cell (10 × 10 mm) was used in all experiments. Argon-saturated acetonitrile solutions containing dicumyl peroxide (1.0 M) were employed. All the experiments were carried out at T = 25 ± 0.5 °C under magnetic stirring. The observed rate constants (kobs) were obtained by averaging two to five individual values and were reproducible to within 5%. The second-order rate constants for the reactions of the cumyloxyl radical with the different substrates were obtained from the slopes of the kobs (measured following the decay of the cumyloxyl radical visible absorption band at 490 nm) vs [substrate] plots.

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.joc.7b02654. A chart showing the substrates and plots of kobs vs substrate concentration for the reactions of CumO• (PDF)

■

AUTHOR INFORMATION

Corresponding Authors

*M.C. e-mail: [email protected]. *M.B. e-mail: [email protected]. ORCID

Miquel Costas: 0000-0001-6326-8299 Massimo Bietti: 0000-0001-5880-7614 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS M.C. acknowledges financial support from MINECO of Spain (CTQ2015-70795-P) and the Catalan DIUE of the Generalitat de Catalunya (2009SGR637). M.C. is thankful for an ICREAAcademia award. We also thank Prof. Lorenzo Stella for the use of LFP equipment. 13548

DOI: 10.1021/acs.joc.7b02654 J. Org. Chem. 2017, 82, 13542−13549

Article

The Journal of Organic Chemistry

■

(27) A 1.5- and 2.2-fold increase in kH for HAT from the C−H bonds of cyclohexane to CumO• was measured on going from isooctane to tert-butyl alcohol and methanol, respectively, a behavior that was explained in terms of solvent hydrogen bonding to the cumyloxyl oxygen atom.28 Along this line, the relatively low concentrations of 1pentanol and pentylamine employed in the kinetic experiments and the linearity of the kobs vs [substrate] plots (see Figure 1) suggest a minor or negligible role for hydrogen bonding from the OH or NH2 group of these substrates to CumO• as an additional activating mechanism. (28) Bietti, M.; Martella, R.; Salamone, M. Org. Lett. 2011, 13, 6110− 6113. (29) By assuming that the three CH2 groups of pentane display a comparable reactivity toward CumO•. This assumption is wellsupported by the results of a recent study on the reaction of CumO• with pentane, carried out in the neat hydrocarbon at T = 160 °C, where a 1.8 ratio between the products deriving from HAT from C2 and C3 was observed.30 (30) Garrett, G. E.; Mueller, E.; Pratt, D. A.; Parent, J. S. Macromolecules 2014, 47, 544−551. (31) Luo, Y.-R. Comprehensive Handbook of Chemical Bond Energies; CRC Press: Boca Raton, FL, 2007. (32) Salamone, M.; DiLabio, G. A.; Bietti, M. J. Am. Chem. Soc. 2011, 133, 16625−16634. (33) González-Núñez, M. E.; Castellano, G.; Andreu, C.; Royo, J.; Báguena, M.; Mello, R.; Asensio, G. J. Am. Chem. Soc. 2001, 123, 7487−7491. (34) Milan, M.; Bietti, M.; Costas, M. ACS Cent. Sci. 2017, 3, 196− 204. (35) Shen, D.; Miao, C.; Wang, S.; Xia, C.; Sun, W. Org. Lett. 2014, 16, 1108−1111. (36) England, P. A.; Rouch, D. A.; Westlake, A. C. G.; Bell, S. G.; Nickerson, D. P.; Webberley, M.; Flitsch, S. L.; Wong, L.-L. Chem. Commun. 1996, 357−358. (37) Salamone, M.; Ortega, V. B.; Bietti, M. J. Org. Chem. 2015, 80, 4710−4715. (38) For the sake of simplicity, only the primary oxidation products observed in these reactions are displayed in Scheme 4. (39) Evidence for rate-determining HAT in metal−oxo-catalyzed C− H oxidations has been supported by the observation of sizable deuterium kinetic isotope effects (KIEs),40−42 of the formation of intermediate carbon-centered radicals,40 and of linear log kH vs C−H BDE correlations.41−43 (40) (a) Liu, W.; Groves, J. T. Acc. Chem. Res. 2015, 48, 1727−1735. (b) Liu, W.; Huang, X. Y.; Cheng, M. J.; Nielsen, R. J.; Goddard, W. A.; Groves, J. T. Science 2012, 337, 1322−1325. (41) (a) Kleespies, S. T.; Oloo, W. N.; Mukherjee, A.; Que, L., Jr. Inorg. Chem. 2015, 54, 5053−5064. (b) Biswas, A. N.; Puri, M.; Meier, K. K.; Oloo, W. N.; Rohde, G. T.; Bominaar, E. L.; Münck, E.; Que, L., Jr. J. Am. Chem. Soc. 2015, 137, 2428−2431. (42) (a) Kwon, E.; Cho, K.-B.; Hong, S.; Nam, W. Chem. Commun. 2014, 50, 5572−5575. (b) Wu, X.; Seo, M. S.; Davis, K. M.; Lee, Y.M.; Chen, J.; Cho, K.-B.; Pushkar, Y. N.; Nam, W. J. Am. Chem. Soc. 2011, 133, 20088−20091. (43) Xue, X.-S.; Ji, P.; Zhou, B.; Cheng, J.-P. Chem. Rev. 2017, 117, 8622−8648. (44) Ramaswamy, S.; Scholze, M.; Plapp, B. V. Biochemistry 1997, 36, 3522−3527. (45) Konishi, H.; Nagase, H.; Manabe, K. Chem. Commun. 2015, 51, 1854−1857.

REFERENCES

(1) Davies, H. M. L.; Morton, D. ACS Cent. Sci. 2017, 3, 936−943. (2) (a) Hartwig, J. F.; Larsen, M. A. ACS Cent. Sci. 2016, 2, 281−292. (b) Hartwig, J. F. J. Am. Chem. Soc. 2016, 138, 2−24. (3) Cernak, T.; Dykstra, K. D.; Tyagarajan, S.; Vachal, P.; Krska, S. W. Chem. Soc. Rev. 2016, 45, 546−576. (4) (a) Brückl, T.; Baxter, R. D.; Ishihara, Y.; Baran, P. S. Acc. Chem. Res. 2012, 45, 826−839. (b) Gutekunst, W. R.; Baran, P. S. Chem. Soc. Rev. 2011, 40, 1976−1991. (5) Newhouse, T.; Baran, P. S. Angew. Chem., Int. Ed. 2011, 50, 3362−3374. (6) White, M. C. Science 2012, 335, 807−809. (7) (a) Salamone, M.; Carboni, G.; Bietti, M. J. Org. Chem. 2016, 81, 9269−9278. (b) Salamone, M.; Bietti, M. Acc. Chem. Res. 2015, 48, 2895−2903. (8) Mack, J. B. C.; Gipson, J. D.; Du Bois, J.; Sigman, M. S. J. Am. Chem. Soc. 2017, 139, 9503−9506. (9) (a) Lee, M.; Sanford, M. S. Org. Lett. 2017, 19, 572−575. (b) Mbofana, C. T.; Chong, E.; Lawniczak, J.; Sanford, M. S. Org. Lett. 2016, 18, 4258−4261. (10) Howell, J. M.; Feng, K.; Clark, J. R.; Trzepkowski, L. J.; White, M. C. J. Am. Chem. Soc. 2015, 137, 14590−14593. (11) Adams, A. M.; Du Bois, J.; Malik, H. A. Org. Lett. 2015, 17, 6066−6069. (12) Asensio, G.; González-Núñez, M. E.; Bernardini, C. B.; Mello, R.; Adam, W. J. Am. Chem. Soc. 1993, 115, 7250−7253. (13) Roberts, B. P. Chem. Soc. Rev. 1999, 28, 25−35. (14) (a) Nanjo, T.; de Lucca, E. C., Jr.; White, M. C. J. Am. Chem. Soc. 2017, 139, 14586−14591. (b) Chen, M. S.; White, M. C. Science 2010, 327, 566−571. (c) Chen, M. S.; White, M. C. Science 2007, 318, 783−787. (15) Milan, M.; Carboni, G.; Salamone, M.; Costas, M.; Bietti, M. ACS Catal. 2017, 7, 5903−5911. (16) Kawamata, Y.; Yan, M.; Liu, Z.; Bao, D.-H.; Chen, J.; Starr, J. T.; Baran, P. S. J. Am. Chem. Soc. 2017, 139, 7448−7451. (17) Yamada, K.; Fukuyama, T.; Fujii, S.; Ravelli, D.; Fagnoni, M.; Ryu, I. Chem. - Eur. J. 2017, 23, 8615−8618. (18) (a) Czaplyski, W. L.; Na, C. G.; Alexanian, E. J. J. Am. Chem. Soc. 2016, 138, 13854−13857. (b) Quinn, R. K.; Könst, Z. A.; Michalak, S. E.; Schmidt, Y.; Szklarski, A. R.; Flores, A. R.; Nam, S.; Horne, D. A.; Vanderwal, C. D.; Alexanian, E. J. J. Am. Chem. Soc. 2016, 138, 696− 702. (c) Schmidt, V. A.; Quinn, R. K.; Brusoe, A. T.; Alexanian, E. J. J. Am. Chem. Soc. 2014, 136, 14389−14392. (19) Wang, D.; Shuler, W. G.; Pierce, C. J.; Hilinski, M. K. Org. Lett. 2016, 18, 3826−3829. (20) Sharma, A.; Hartwig, J. F. Nature 2015, 517, 600−604. (21) Jeffrey, J. L.; Terrett, J. A.; MacMillan, D. W. C. Science 2015, 349, 1532−1536. (22) (a) Canta, M.; Font, D.; Gómez, L.; Ribas, X.; Costas, M. Adv. Synth. Catal. 2014, 356, 818−830. (b) Prat, I.; Gómez, L.; Canta, M.; Ribas, X.; Costas, M. Chem. - Eur. J. 2013, 19, 1908−1913. (c) Gómez, L.; Canta, M.; Font, D.; Prat, I.; Ribas, X.; Costas, M. J. Org. Chem. 2013, 78, 1421−1433. (d) Gómez, L.; Garcia-Bosch, I.; Company, A.; Benet-Buchholz, J.; Polo, A.; Sala, X.; Ribas, X.; Costas, M. Angew. Chem., Int. Ed. 2009, 48, 5720−5723. (23) (a) Moteki, S. A.; Selvakumar, S.; Zhang, T.; Usui, A.; Maruoka, K. Asian J. Org. Chem. 2014, 3, 932−935. (b) Moteki, S. A.; Usui, A.; Zhang, T.; Solorio Alvarado, C. R.; Maruoka, K. Angew. Chem., Int. Ed. 2013, 52, 8657−8660. (24) Ottenbacher, R. V.; Samsonenko, D. G.; Talsi, E. P.; Bryliakov, K. P. Org. Lett. 2012, 14, 4310−4313. (25) (a) Baciocchi, E.; Bietti, M.; Salamone, M.; Steenken, S. J. Org. Chem. 2002, 67, 2266−2270. (b) Avila, D. V.; Ingold, K. U.; Di Nardo, A. A.; Zerbetto, F.; Zgierski, M. Z.; Lusztyk, J. J. Am. Chem. Soc. 1995, 117, 2711−2718. (26) (a) Weber, M.; Fischer, H. J. Am. Chem. Soc. 1999, 121, 7381− 7388. (b) Salamone, M.; Basili, F.; Mele, R.; Cianfanelli, M.; Bietti, M. Org. Lett. 2014, 16, 6444−6447. 13549

DOI: 10.1021/acs.joc.7b02654 J. Org. Chem. 2017, 82, 13542−13549