Article pubs.acs.org/joc
Resonance Destabilization in N‑Acylanilines (Anilides): ElectronicallyActivated Planar Amides of Relevance in N−C(O) Cross-Coupling Roman Szostak,‡ Guangrong Meng,† and Michal Szostak*,† ‡
Department of Chemistry, Wroclaw University, F. Joliot-Curie 14, Wroclaw 50-383, Poland Department of Chemistry, Rutgers University, 73 Warren Street, Newark, New Jersey 07102, United States
†
S Supporting Information *
ABSTRACT: Transition-metal-catalyzed activation of amide N−C(O) bonds proceeds via selective metal insertion into the carbon−nitrogen amide bond. Herein, we demonstrate that Nacylanilines (anilides), the first class of planar amides that have been shown to undergo selective amide N−C cross-coupling reactions, feature a significantly decreased barrier to rotation around the amide N−C(O) bond. Most significantly, we demonstrate that amide nN → π*CO resonance in simple anilides can be varied by as much as 10 kcal/mol. The data have important implications for the design of N−C(O) amide cross-coupling reactions and control of the molecular conformation of anilides by resonance effects.
1. INTRODUCTION Amidic resonance is one of the most fundamental effects in organic chemistry that has been extensively utilized to control conformation and reactivity of atoms comprising the amide linkage.1−3 The hindered barrier to rotation around the N− C(O) bond determines the overall structure of molecules featuring amides.4,5 This effect has been utilized in molecular recognition, conformational relays, and drug−receptor interactions.1−5 Amidic resonance in simple N-acylanilines (anilides) has long been recognized as a key feature controlling cis− trans conformational preference of the amide bond.6 Selective transition-metal-catalyzed functionalization of N− C(O) bonds in amides by metal insertion into the N−C bond has emerged as a valuable method in organic synthesis (Figure 1A).7−9 Success in this field is expected to open myriad perspectives for the cross-coupling of bench-stable, ubiquitous amide building blocks with a greatly expanded scope over other carboxylic acid electrophiles via acyl and aryl cross-coupling mechanisms under redox neutral conditions.10,11 Extensive studies have demonstrated that the amide bond reactivity in metal-catalyzed N−C functionalization can be correlated with the extent of amidic resonance.8g Therefore, it is critical that it be possible to rationally modulate resonance in a broad range of amide cross-coupling partners. Despite the emergence of new methods, activation of the amide carbon−nitrogen bond in anilides is rare. Following the breakthrough report by Garg and co-workers on Ni-catalyzed esterification of anilides (Figure 1B),8a these precursors have been typically considered inert in an array of N−C amide bond functionalizations.7−9 To understand the role of amidic resonance on the reactivity of anilides in N−C cross-coupling manifolds, we examined resonance energies in a series of anilides to determine their effect on the amide N−C(O) bond reactivity (Figure 1C). (i) To date, amide bond resonance © 2017 American Chemical Society
Figure 1. (a) Amide N−C bond activation. (b) Anilides in N−C bond cross-coupling. (c) Electronic effects in anilides. ER = resonance energy; lp = lone pair.
energies in N-acylanilides (anilides) of relevance to N−C crosscoupling have not been determined.12 (ii) The goal of the study was to determine amidic resonance in N-acylanilides (anilides). It is well established that anilides and related compounds show cis-conformational preference of the amide bond.6i (iii) The resonance has been estimated using the Carbonyl Substitution Nitrogen Atom Replacement (COSNAR) method.13 This Received: April 23, 2017 Published: June 7, 2017 6373
DOI: 10.1021/acs.joc.7b00971 J. Org. Chem. 2017, 82, 6373−6378
Article
The Journal of Organic Chemistry method represents one of the most reliable methods for the estimation of amidic resonance.13a−e (iv) The COSNAR method provides calculated structures and energies of the corresponding amines, ketones, and hydrocarbons, all of which are available for examination of structural changes that occur during steric and/or electronic variations of the amide bond in comparison with amide bond analogues. (v) We have performed extensive method optimization and determined N/ O-protonation, aptitude and rotational profile in a representative anilide in the class. (vi) The presented data demonstrate that simple substitution of aromatic rings allows one to change amidic resonance in anilides to values less than 10 kcal/mol (ca. 50% of the amide bond amidicity), which is critical for the amide N−C cross-coupling.7,8g Furthermore, the present finding pertaining to amidic resonance in anilides may also find utility in metal-free nucleophilic additions to the amide bond.7a,b The data presented show several important features: (1) Anilides, the first class of planar amides that have been shown to undergo selective amide N−C cross-coupling reactions, feature a significantly decreased barrier to rotation around the amide N−C(O) bond.8a (2) Electronic effects can be applied to decrease amidic resonance to 70% change in amidicity).3d The data have important implications for the design of new N−C(O) amide cross-coupling reactions and the control of molecular conformation of anilides by electronic effects.
Figure 2. Structures of amides employed in the present study.
Table 1. Resonance Energies for Anilide N−C(O) Rotation Calculated using B3LYP/6-311++G(d,p)a
entry
R
amide
ER [kcal/mol]
σ
σ+
1 2 3 4 5 6 7 8 9 10 11 12
4−H 4−MeO 4−CN 4−F 4−Br 4−NO2 4−CF3 4−Cl 4−Me 4−NMe2 4−COMe 3−MeO
1a 1b 1c 1d 1e 1f 1g 1h 1i 1j 1k 1l
13.5 12.8 14.7 13.6 13.9 15.1 14.5 13.8 13.2 11.9 14.4 13.5
0 −0.27 0.66 0.06 0.23 0.78 0.54 0.23 −0.17 −0.83 0.50 0.12
0 −0.78 0.66 −0.07 0.15 0.79 0.61 0.11 −0.31 −1.70 0.49 0.05
a
Energies and bond lengths, see Table SI-2. Representative data on acyclic twisted amides: refs 12a, b. Representative data on bridged lactams: refs 13, 14.
2. RESULTS AND DISCUSSION The COSNAR method was used to calculate the resonance energies of the anilides selected for the study (eq 1, ET = total energy).13a,b −RE = E T(amide) − [E T(amine) + E T(ketone) − E T(hydrocarbon)] (1)
Extensive method optimization was performed using DFT and ab initio methods (Table SI-1, Supporting Information, SI). B3LYP/6-311++G(d,p) was selected to conduct geometry optimization of amides, amines, ketones, and hydrocarbons as a result of the good reproducibility of literature data6b and the method practicality (note that the method was verified by obtaining a good correlation with the available literature data: 1a, ER = 13.3 ± 0.3 kcal/mol, lit; 13.5 kcal/mol, B3LYP/6-311+ +G(d,p)). To obtain information on the effect of electronic stabilization of the substituent on the aromatic ring, a series of anilides 1 featuring various electron-rich (NMe2, OMe, Me), neutral (H), electron-withdrawing (NO2, CN, COMe, CF3), and halide (Br, Cl, F) substituents arranged according to the potential utility in cross-coupling reactions were examined (Figure 2, Table 1). Examination of the data in Table 1 reveals that changing from 4−NMe2 (entry 10, 1j, ER = 11.9 kcal/mol) to 4−NO2 (entry 6, 1f, ER = 15.1 kcal/mol) leads to a ΔER of 3.2 kcal/mol. This corresponds to a 2 × 102-fold decrease in the rate of anilide N− C(O) rotation at 298 K.15 The Hammett correlation study (ER vs σ) showed a large positive ρ-value of 1.95 (R2 = 0.99) (Figure 3), which can be
Figure 3. Plot of ER (kcal/mol) vs σ for anilide N−C(O) rotation: 1.
compared with the ρ+ value of 1.23 (R2 = 0.95) using Hammett−Brown σ+ constants, indicating that rotation around the N−C(O) bond involves a decrease in nN → π*CO conjugation in 1. The following changes in bond lengths support an increase of N−C(O) rotation for electron-rich substitution of the aromatic ring: (1) shortening of the C−C bond; (2) elongation of N−C(O) and CO bonds; (3) shortening of the N−C(Ar) bond (see Table SI-2 in the SI for 6374
DOI: 10.1021/acs.joc.7b00971 J. Org. Chem. 2017, 82, 6373−6378
Article
The Journal of Organic Chemistry
rotation. This effect is supported by the following changes in bond lengths: (1) shortening of the N−C(Ar) bond; (2) elongation of the N−C(O) and shortening of the CO bond; (3) shortening of the C−C(O) bond (see Table SI-2 in the SI for bond lengths of amides 2). Overall, these structural and electronic changes are consistent with a strong nN → Ar conjugation.16 The additive effect of Hammett parameters on the barrier of anilide N−C(O) rotation is presented in Scheme 1.17 By
bond lengths of amides 1). Collectively, these structural and electronic changes in 1 strongly suggest that (1) resonance tuning by electronic effects on the aromatic ring, and (2) Ar1− C(O) to N−Ar2 conformational communication in anilides 1 are feasible. The effect of the N-aryl ring on electronic destabilization was explored (Table 2). Remarkably, the data reveal that changing Table 2. Resonance Energies for Anilide N−C(O) Rotation Calculated using B3LYP/6-311++G(d,p)a
Scheme 1. Resonance Energies for N−C(O) Rotation
a
entry
R
amide
ER [kcal/mol]
σ
σ+
1 2 3 4 5 6 7 8 9 10 11 12
4−H 4−MeO 4−CN 4−F 4−Br 4−NO2 4−CF3 4−Cl 4−Me 4−NMe2 4−COMe 3−MeO
1a 2b 2c 2d 2e 2f 2g 2h 2i 2j 2k 2l
13.5 16.2 11.1 14.0 13.2 10.3 11.8 13.3 14.0 16.8 11.3 13.1
0 −0.27 0.66 0.06 0.23 0.78 0.54 0.23 −0.17 −0.83 0.50 0.12
0 −0.78 0.66 −0.07 0.15 0.79 0.61 0.11 −0.31 −1.70 0.49 0.05
changing a single substituent on both of the rings the rate of rotation can be varied by a remarkable ER of almost 10 kcal/ mol (3a vs 3d). This corresponds to >70% change in amidicity (vs 1a) and a 5 × 107-fold change in the rate of rotation at 298 K.15 The substituent effect on the resonance in anilides is summarized in Scheme 2. Note the high magnitude of Scheme 2. Summary of Electronic Effects Favoring Anilide N−C(O) Rotation (Red Arrow Indicates 1° Effect)
See Table 1.
from 4−NO2 (entry 6, 2f, ER = 10.3 kcal/mol) to 4−NMe2 (entry 10, 2i, ER = 16.8 kcal/mol) leads to a much higher ΔER of 6.5 kcal/mol. This corresponds to a 4 × 105-fold decrease in the rate of anilide N−C(O) rotation at 298 K.15 The Hammett plot (ER vs σ) shows a large negative ρ-value of −4.18 (R2 = 0.93, ρ+-value of −2.68, R2 = 0.91) (Figure 4), indicating that anilides containing electron-withdrawing substituents on the N-Ar ring undergo more rapid N−C(O)
destabilization by electron-withdrawing groups on the N-Ar ring. In addition, a combination of two EDG or EWG gives an intermediate value of the resonance. For example, for the most stabilizing/destabilizing substituents (R/R′ = 4−NO2/4−NO2 and R/R′ = 4−NMe2/4−NMe2), the values are as follows: R/ R′ = 4−NO2/4−NO2: 11.3 kcal/mol; R/R′ = 4−NMe2/4− NMe2: 15.3 kcal/mol (not shown), which can be compared with the resonance in the corresponding anilides: R/R′ = 4− NO2/H: 15.1 kcal/mol and R/R′ = H/4−NO2: 10.2 kcal/mol; R/R′ = 4−NMe2/H: 11.9 kcal/mol and R/R′ = H/4−NMe2: 16.8 kcal/mol. The resonance in these EDG/EDG and EWG/ EWG examples is consistent with the electronic effect of the Nsubstituent as the major contributing factor to the resonance in these amides. The effect of both the C-aryl and N-aniline groups on the amidic resonance was determined using isodesmic equations for the most stabilizing/destabilizing substituents (4−NO2 and 4− NMe2) (not shown). We found, as expected from the connectivity, that there is a reasonably good correlation between the changes in the resonance energy of the aniline component (4−NO2: 3.2 kcal/mol; 4−NMe2: − 2.6 kcal/mol, which can be compared with the difference in the corresponding anilides: 4−NO2: 3.2 kcal/mol; 4−NMe2: − 3.3 kcal/mol). However, the C-aryl group does not give a good correlation, likely because of conformational changes of the amide bond that occur during the substitution (4−NO2: − 2.5
Figure 4. Plot of ER (kcal/mol) vs σ for anilide N−C(O) rotation: 2. 6375
DOI: 10.1021/acs.joc.7b00971 J. Org. Chem. 2017, 82, 6373−6378
Article
The Journal of Organic Chemistry
Figure 5. (a) Rotational profile of 1a (ΔE, kcal/mol, vs O−C−N−C [°]). (b) Minimum and maximum conformations. The insets show Newman projections along the N−C(O) axis.
kcal/mol; 4−NMe2: 3.4 kcal/mol, which can be compared with the difference in the corresponding anilides: 4−NO2: − 1.6 kcal/mol; 4−NMe2: 1.6 kcal/mol). The effect of individual functional groups on amidic resonance will be the subject of future studies. To gain additional insight into the most-stable conformation of 1a, the rotational profile of anilide 1a as a model for N−C amide cross-coupling was obtained to determine the relationship between energy and N−C(O) geometry by systematic rotation along the O−C−N−C(Me) dihedral angle (Figure 5A). The rotation was performed in both directions. We employed the structure of N-methyl-N-phenylbenzamide as the starting geometry and performed full optimization.6c The rotational profile is symmetrical, with the trans-conformer being less stable than the cis-conformer by 3.2 kcal/mol. Most importantly, the profile confirms the electronic destabilization of 1a (cf. twist). The energy minimum is located at approximately a 0° O−C−N−C dihedral angle. The energy maximum corresponds to approximately a 70° dihedral angle (τ = 88.4°). There is a second energy minimum at approximately a 170° dihedral angle (anti N−Me/CO conformation). Minimum and maximum conformations of 1a as well as Newman projections along the N−C(O) axis are shown in Figure 5B. It is worth noting that the difference between N- and Oprotonation (protonation affinities, PA) in representative 1a indicate that these anilides vastly favor protonation at oxygen (ΔPA = 11.0 kcal/mol).13,14 Thus, activation of the acyl group by N-protonation in these compounds is unlikely. The generalized reactivity scale for the amide N−C(O) bond activation is presented in Scheme 3. The reactive N-methyl-Nphenylbenzamide (1a, ER = 13.5 kcal/mol) features a
Scheme 3. Reactivity Scale of Anilides Relevant to MetalCatalyzed N−C Cross-Couplinga
a Reactivity of anilides in N−C insertion is rare.8a The majority of amides undergoing N−C coupling feature lower ER than anilides.7−9
substantially lower ER than unreactive N,N-dimethylbenzamide (1a′, ER = 16.0 kcal/mol) and N,N-dimethylacetamide (1a″, ER = 18.3 kcal/mol).7−9 Resonance energies in the α-alkyl anilide series 4 are also much lower than in N,N-dimethylacetamide (18.3 kcal/mol): 4a, 15.4 kcal/mol; 4b, 14.9 kcal/mol; 4c, 15.2 kcal/mol; 4d, 9.9 kcal/mol (Figure 2). Note that (1) activation of the amide bond in anilides sets the limits for the current synthetically useful activation of N−C bonds, and (2) the barrier of approximately 10 kcal/mol sets the limits for the current Pd-catalyzed amide N−C(O) crosscoupling.7−9 We fully expect that rational resonance fine-tuning will enable synthetically useful selectivity in amide N−C(O) cross-coupling. 6376
DOI: 10.1021/acs.joc.7b00971 J. Org. Chem. 2017, 82, 6373−6378
Article
The Journal of Organic Chemistry
(4) Selected studies using bridged lactams: (a) Tani, K.; Stoltz, B. M. Nature 2006, 441, 731. (b) Liniger, M.; VanderVelde, D. G.; Takase, M. K.; Shahgholi, M.; Stoltz, B. M. J. Am. Chem. Soc. 2016, 138, 969. (c) Komarov, I. V.; Yanik, S.; Ishchenko, A. Y.; Davies, J. E.; Goodman, J. M.; Kirby, A. J. J. Am. Chem. Soc. 2015, 137, 926. (d) Sliter, B.; Morgan, J.; Greenberg, A. J. Org. Chem. 2011, 76, 2770. (e) Artacho, J.; Ascic, E.; Rantanen, T.; Karlsson, J.; Wallentin, C. J.; Wang, R.; Wendt, O. F.; Harmata, M.; Snieckus, V.; Wärnmark, K. Chem. - Eur. J. 2012, 18, 1038. (5) Selected studies using acyclic amides: (a) Clayden, J.; Lund, A.; Vallverdu, L.; Helliwell, M. Nature 2004, 431, 966. (b) Clayden, J. Chem. Soc. Rev. 2009, 38, 817. (c) Sola, J.; Fletcher, S. P.; Castellanos, A.; Clayden, J. Angew. Chem., Int. Ed. 2010, 49, 6836. (d) Knipe, P. C.; Thompson, S.; Hamilton, A. D. Chem. Sci. 2015, 6, 1630. (e) Barrett, K. T.; Metrano, A. J.; Rablen, P. R.; Miller, S. J. Nature 2014, 509, 71. (f) Shah, N. H.; Butterfoss, G. L.; Nguyen, K.; Yoo, B.; Bonneau, R.; Rabenstein, D. L.; Kirshenbaum, K. J. Am. Chem. Soc. 2008, 130, 16622 and references cited therein. (6) (a) Itai, A.; Toriumi, Y.; Tomioka, N.; Kagechika, H.; Azumaya, I.; Shudo, K. Tetrahedron Lett. 1989, 30, 6177. (b) Azumaya, I.; Kagechika, H.; Yamaguchi, K.; Shudo, K. Tetrahedron 1995, 51, 5277. (c) Itai, A.; Toriumi, Y.; Saito, S.; Kagechika, H.; Shudo, K. J. Am. Chem. Soc. 1992, 114, 10649. (d) Yamaguchi, K.; Matsumura, G.; Kagechika, H.; Azumaya, I.; Ito, Y.; Itai, A.; Shudo, K. J. Am. Chem. Soc. 1991, 113, 5474. (e) Azumaya, I.; Yamaguchi, K.; Okamoto, I.; Kagechika, H.; Shudo, K. J. Am. Chem. Soc. 1995, 117, 9083. (f) Yamasaki, R.; Tanatani, A.; Azumaya, I.; Saito, S.; Yamaguchi, K.; Kagechika, H. Org. Lett. 2003, 5, 1265. (g) Okamoto, I.; Nabeta, M.; Hayakawa, Y.; Morita, N.; Takeya, T.; Masu, H.; Azumaya, I.; Tamura, O. J. Am. Chem. Soc. 2007, 129, 1892. (h) Katoono, R.; Tanaka, Y.; Fujiwara, K.; Suzuki, T. J. Org. Chem. 2014, 79, 10218. (i) The cisconformational preference of N-alkylanilides is well established. See, refs 5i, 6a−c. For additional studies, see: (j) Zhang, X.; Sun, X. Y.; Wang, C. J.; Jiang, Y. B. J. Phys. Chem. A 2002, 106, 5577. (k) Broxton, T. J.; Deady, L. W.; Pang, Y. T. Tetrahedron Lett. 1975, 16, 2799. (7) Reviews on N−C amide cross-coupling: (a) Meng, G.; Shi, S.; Szostak, M. Synlett 2016, 27, 2530. (b) Liu, C.; Szostak, M. Chem. Eur. J. 2017, 23, 7157. (c) Dander, J. E.; Garg, N. K. ACS Catal. 2017, 7, 1413. (8) For representative examples of acyl coupling of amides, see: (a) Hie, L.; Fine Nathel, N. F.; Shah, T. K.; Baker, E. L.; Hong, X.; Yang, Y. F.; Liu, P.; Houk, K. N.; Garg, N. K. Nature 2015, 524, 79. (b) Weires, N. A.; Baker, E. L.; Garg, N. K. Nat. Chem. 2015, 8, 75. (c) Baker, E. L.; Yamano, M. M.; Zhou, Y.; Anthony, S. M.; Garg, N. K. Nat. Commun. 2016, 7, 11554. (d) Hie, L.; Baker, E. L.; Anthony, S. M.; Desrosiers, J. N.; Senanayake, C.; Garg, N. K. Angew. Chem., Int. Ed. 2016, 55, 15129 and references cited therein. (e) Meng, G.; Szostak, M. Org. Lett. 2015, 17, 4364. (f) Meng, G.; Shi, S.; Szostak, M. ACS Catal. 2016, 6, 7335. (g) Lei, P.; Meng, G.; Szostak, M. ACS Catal. 2017, 7, 1960 and references cited therein. (9) For representative examples of decarbonylative coupling of amides, see: (a) Meng, G.; Szostak, M. Angew. Chem., Int. Ed. 2015, 54, 14518. (b) Shi, S.; Meng, G.; Szostak, M. Angew. Chem., Int. Ed. 2016, 55, 6959. (c) Meng, G.; Szostak, M. Org. Lett. 2016, 18, 796. (d) Hu, J.; Zhao, Y.; Liu, J.; Zhang, Y.; Shi, Z. Angew. Chem., Int. Ed. 2016, 55, 8718. (e) Dey, A.; Sasmal, S.; Seth, K.; Lahiri, G. K.; Maiti, D. ACS Catal. 2017, 7, 433. (f) Yue, H.; Guo, L.; Lee, S. C.; Liu, X.; Rueping, M. Angew. Chem., Int. Ed. 2017, 56, 3972. (g) Hu, J.; Wang, M.; Pu, X.; Shi, Z. Nat. Commun. 2017, 8, 14993. (10) (a) Metal-Catalyzed Cross-Coupling Reactions and More; de Meijere, A., Bräse, S., Oestreich, M., Eds.; Wiley: New York, 2014. (b) Science of Synthesis: Cross Coupling and Heck-Type Reactions; Molander, G. A., Wolfe, J. P., Larhed, M., Eds.; Thieme: Stuttgart, 2013. (c) Johansson-Seechurn, C. C. C.; Kitching, M. O.; Colacot, T. J.; Snieckus, V. Angew. Chem., Int. Ed. 2012, 51, 5062. (11) Review on acyl-metal intermediates: Gooßen, L. J.; Rodriguez, N.; Gooßen, K. Angew. Chem., Int. Ed. 2008, 47, 3100. (12) For mechanistic studies on N−C bond cleavage, see: (a) Szostak, R.; Shi, S.; Meng, G.; Lalancette, R.; Szostak, M. J. Org. Chem. 2016,
3. CONCLUSIONS In summary, the results presented in this paper can have a major effect on the design of selective N−C cross-coupling reactions. We have demonstrated that N-acylanilines (anilides), the first class of planar amides to participate in selective amide N−C cross-coupling reactions, feature a significantly decreased barrier to rotation around the N−C(O) bond. The barrier to rotation in anilides can be varied by values approaching 10 kcal/mol by simple electronic fine-tuning. Electronic effects can be applied to decrease amidic resonance to
