Mechanistic Insights into Solvent and Ligand Dependency in Cu(I

Subscriber access provided by ECU Libraries

Article

Mechanistic Insights into the Solvent and Ligand-Dependency in Cu(I)-Catalyzed Allylic Alkylation with gem-Diborylalkanes Qi Zhang, Bing Wang, Jia-Qin Liu, Yao Fu, and Yu Cheng Wu J. Org. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.joc.7b02249 • Publication Date (Web): 15 Dec 2017 Downloaded from http://pubs.acs.org on December 15, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Organic Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Organic Chemistry

Mechanistic Insights into the Solvent and Ligand-Dependency in Cu(I)-Catalyzed Allylic Alkylation with gem-Diborylalkanes

Qi Zhang, a,c Bing Wang, b Jia-Qin Liu, *a,c Yao Fu, *b Yu-Cheng Wua,c a

Institute of Industry & Equipment Technology, Hefei University of Technology, Hefei 230009 b

c

Department of Chemistry, University of Science and Technology of China, Hefei 230026

Key Laboratory of Advanced Functional Materials and Devices of Anhui Province, Hefei 230009, China

Emails: [email protected]; [email protected]

Table of Content

1

ACS Paragon Plus Environment

The Journal of Organic Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract The recent Cu-catalyzed allylic substitution reaction between gem-diboryalkane and allyl electrophiles shows intriguing solvent and ligand-controlled regioselectivity. The α-alkylation product was obtained in DMF solvent, while γ-alkylation product was obtained in dioxane solvent and the dioxane & NHC ligand situation. In the present study, density functional theory (DFT) calculations have been used to investigate the reaction mechanism and origin of the regioselectivity. For both of dioxane and DMF, γ-alkylation undergoes successive oxidative addition (CH2Bpin trans to leaving group) and direct Cγ-C reductive elimination. The α-alkylation is found to undergo oxidative addition (CH2Bpin trans to leaving group), isomerization and Cα-C reductive elimination rather than the previously proposed oxidative addition (-CH2Bpin cis to the leaving group) and Cα-C reductive elimination. The γ-alkylation and α-alkylation is respectively favorable for dioxane and DMF solvent, which is consistent with the γand α-selectivity in experiment. The solvent interferes the isomerization step, thereby affects the relative facility of the α- and γ-alkylation. Further investigation shows that η1-intermediate formation promoted by solvent is the rate-determining step of the isomerization. The stronger electron-donating ability of DMF than dioxane facilitates the η1-intermediate formation and finally results in the easier isomerization in DMF. For dioxane & NHC situation, in the presence of neutral NHC ligand, the -PO4Et2 group tightly coordinates with the Cu center after the oxidative addition, preventing the isomerization process. The regioselectivity is determined by the relative facility of the oxidative addition step. Therefore, the favorable oxidative addition (in which -CH2Bpin trans to the leaving group) results in the facility of γ-alkylation.

2


Page 2 of 24

Page 3 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


1. Introduction The Suzuki-Miyaura coupling (SMC) reaction with alkylboron acting as nucleophile provides a mild and efficient method to construct C(sp3)-C bonds in synthesizing complex organic compounds.1 Among these, the gem-diboryalkane is a novel and readily available alkylboron nucleophile containing two boron atoms.2 One gem-boron could assist the transmetallation of another one, and the retaining boron could participate in further coupling reaction to obtain complex multi-substituted compounds.3 For the first time, the Shibata group4 reported the Pd-catalyzed SMC reaction between gem-diboryalkane and aryl halide (Scheme 1a). Through this reaction,

various

multi-substituted

C(sp3)-C(sp2)

bonds

were

successfully

constructed.5 Additionally, benzyl and allyl electrophiles could also react with gem-diboryalkane under the similar Pd-catalyzed condition.6 Later, by using chiral phosphine ligand with Pd catalyst, the Morken group accomplished the stereoselective SMC between gem-diboryalkane and aryl7a and alkenyl7b electrophiles (Scheme 1b). They also achieved the non-catalytic reaction between gem-diboryalkane and primary alkyl electrophiles to construct C(sp3)-C(sp3) bond (Scheme 1c).8 The same reaction was also reported by our group via Cu-catalyzed system (the allyl electrophile was also tried).9 Recently, using the [NHC]Cu-catalyzed system (NHC refers to N-heterocyclic carbene), Cho10 and our group11 independently reported the coupling between gem-diboryalkane and allyl electrophiles (Scheme 1d).

3



Page 4 of 24

Scheme 1. The gem-diboryalkane participated alkylation reaction

In the allyl electrophiles participated alkylation reactions reported by our group (Scheme 1c & 1d), regioselectivity could be successfully regulated by solvent and ligand.

9, 11

As shown in Scheme 2a, with CuCl as catalyst, MeOLi as base, the

γ-alkylation products were yielded with dioxane as solvent. The γ-selectivity further increased (from 83:17 to 97:3) while NHC ligand was added. However, by changing solvent dioxane to DMF in the absence of NHC, α-alkylation products were surprisingly yielded. The intriguing solvent and ligand-controlled regioselectivity motivates us to investigate the mechanism of the Cu-catalyzed allyl substitution reaction. According to the studies of Nakamura,12 Bäckvall,13 Ogle14 et al, the oxidative

addition-reductive

elimination

mechanism

via

successive

alkene

coordination, SN2'-type oxidative addition and reductive elimination is proposed (Scheme 2b). Meanwhile, the alkene insertion-β-elimination mechanism via successive alkene coordination, alkene insertion and β-elimination is also possible based on the proposals of Sawamura and Ohmiya.15 Therefore, the mechanism of the Cu-catalyzed allyl substitution reaction is unclear. More importantly, the roles of the 4


Page 5 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


solvent and NHC ligand controlling the α/γ-regioselectivity are not reflected in the mechanism proposals. The mechanistic origin for the regulated regioselectivity is worth clarification.

Scheme 2. (a) The solvent and ligand-controlled regioselectivity, (b) The mechanism proposal of Cu-catalyzed allyl substitution reaction In order to solve these problems, we carried out theoretical analysis on the mechanism of the Cu-catalyzed allyl alkylation reaction (Scheme 2a). Three reaction conditions of solvent dioxane, solvent DMF and dioxane solvent-NHC ligand were investigated. The calculation results for the mechanism of solvent dioxane and DMF are shown in Section 3.1 and Sections 3.2, and Sections 3.3 shows the comparison of these two conditions. The calculation results for the mechanism of dioxane solvent-NHC ligand situation is shown in Sections 3.4.

2. Computational Methods and Model Reaction 2.1 Computational Methods

5



All calculations in this study were carried out in Gaussian 09 program.16 The B3LYP17,18/GEN1 method (GEN1: 6-31g* for C, H, O, N, P, B, Li and LANL2DZ for Cu) combined with SMD model19 was used for geometry optimization in solvent DMF and dioxane (consistent with our experiments9,11). To gain the thermodynamic corrections of Gibbs free energy and verify the stationary points to be local minima or saddle points, we conducted frequency analysis at the same level with optimization (zero imaginary frequency for local minima and one for saddle point). For all transition states, we performed the intrinsic reaction coordinate (IRC) analysis to confirm that they connect the correct reactants and products on the potential energy surface.20 The M0621/GEN2 method (GEN2: 6-311++G** for C, H, O, N, P, B, Li and SDD for Cu) method with the SMD model was used for the solution-phase single-point energy calculations of all of these stationary points. The polarization function was added for Cu(ζ(f) = 3.525).22,23 All energetics involved in this study are calculated by adding the Gibbs free energy correction calculated at B3LYP/GEN1 and the single-point energy calculated with the M06/GEN2 method.24 2.2 Model Reaction

Scheme 3 The model reaction The reaction of allyl electrophile 1a and gem-diboryalkane 2a generating γ-alkylation product 3a and α-alkylation product 4a is chosen as the model reaction (Scheme 3). CuCl and MeOLi are used as the catalyst and base, respectively. 3. Results and Discussions 3.1 Mechanism with solvent dioxane

6


Page 6 of 24

Page 7 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The mechanism for the situation of solvent dioxane is firstly investigated. In this section,

both

oxidative

addition-reductive

elimination

and

alkene

insertion-β-elimination mechanisms are examined to determine the favorable mechanism. Before the reaction, the original catalyst CuCl reacts with the base MeOLi and diborylmethane 2a to generate cat1 (Li(BpinCH2CuOMe)) with energy decrease of 50.6 kcal/mol (Figure 1). Ion separation of cat1 generates anionic Cu(I) species cat2 (BpinCH2CuOMe-) with energy increase of 50.5 kcal/mol. Dimerization of cat1 generates cat3 with energy decrease of 20.9 kcal/mol. 25,26 Considering the lower free energy than that of cat1 and cat2, the following steps were investigated from cat3.

Figure 1. The different forms of Cu(I) catalyst Oxidative addition-reductive elimination mechanism Efforts were first put into examining the oxidative addition-reductive elimination mechanism, in which alkene coordination, SN2'-type oxidative addition and reductive elimination occur successively (Figure 2).12-14 From cat3, the dissociation of one Li-OMe or Li-Bpin interaction is prerequisite to form vacant coordinating site on the Cu(I) center. As shown in Figure 2, break of one Li-Bpin interaction (i.e. Lia-Bpin) allows the coordination of substrate 1a to the Cu center. The η3-intermediate Int1 is generated with the -CH2Bpin group trans to the leaving -PO4Et2 group. New interaction forms between Lia and O of the -PO4Et2 group, additionally. Break of one Li-OMe interaction (i.e. Lia-OMe) and 1a coordination generate η3-intermediate Int2 7



with the -OMe group trans to -PO4Et2. The free energies of Int1 and Int2 are -9.3 and 3.9 kcal/mol, respectively.

Figure 2. The detailed oxidative addition-reductive elimination mechanism in solvent dioxane (Path-α and Path-γ) The SN2′-type oxidative addition then occurs from Int1 and Int2. As shown in Figure 2, Int1 undergoes oxidative addition via transition state TS1, in which CH2Bpin occupies the trans position of the leaving group. Additionally, the Lia+ assists the C-PO4Et2 bond cleavage while the Lib+ cation acts as a bridging between these two BpinCuOMe- moieties. The free energy of TS1 is -2.7 kcal/mol, and the energy barrier is 18.2 kcal/mol (cat3→TS1). The η3-Cu(III) intermediate Int3 was then generated with the decreased free energy of -26.1 kcal/mol. Similarly, Int2 undergoes oxidative addition via TS2, in which OMe occupies the trans position of the leaving group. The free energy of TS2 is 12.8 kcal/mol, and the energy barrier is 33.7 kcal/mol (cat3→TS2). Int4 was generated with free energy of -19.5 kcal/mol. Therefore, the former oxidative addition process is more favorable than the latter, indicating that the CH2Bpin group favorably occupies the trans position of the leaving group. It is understandable because the stronger trans effect of CH2Bpin (than OMe) benefits the C-PO4Et2 bond cleavage. Additionally, we found that the smaller twisting 8


Page 8 of 24

Page 9 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


degree of the bridging structure Lia…BpinCH2CuOMe…Lib also contributes to the facility of the former oxidative addition.27 The generated Int3 and Int4 subsequently dissociate the LiCuLi structure to generate the mononuclear intermediate Int5 and Int6. 12a,28 With the -CH2Bpin group respectively cis to the Cγ atom and Cα atom in these two intermediates, direct Cγ-C and Cα-C reductive elimination occur from them to yield the γ-alkylation product 3a and α-alkylation product 4a. The corresponding transition states are TS3 and TS4, with free energies of -4.6 and -7.0 kcal/mol. Finally, the Cu(I) catalyst CuOMe is regenerated, and the free energy decreases to -29.1 and -32.4 kcal/mol, respectively. As shown in Figure 2, the pathway via TS1 (oxidative addition transition state) and TS3 (reductive elimination transition state) is named as Path-γ (blue line) for γ-alkylation. Meanwhile, the pathway via TS2 (oxidative addition transition state) and TS4 (reductive elimination transition state) is named as Path-α (brown line) for α-alkylation. Both Path-α and Path-γ are irreversible (because the free energies of TS3 and TS4 are lower than that of TS1 and TS2), indicating that the oxidative addition step determines the relative facility of Path-α and Path-γ. Since the oxidative addition via TS1 is favorable than that via TS2, Path-γ is favorable than Path-α, which is consistent with the γ-selectivity for dioxane solvent in experiment. However, the energy barrier difference of these two oxidative addition steps is 15.5 kcal/mol. The large energy barrier difference does not match the experimental phenomenon of small amount of α-product was generated. The inconsistent phenomenon inspires us to find more favorable α-alkylation mechanism.

9



Figure 3. The isomerization process from Int5 to Int6 Considering the solvent-controlled selectivity and the solvent-coordinated model proposed by Yamanaka et al.,12b we suggest that the solvent might participate in the isomerization of two reductive elimination precursors (Int5 & Int6). The α-product might be generated via isomerization (from Int5 to Int6) and subsequent Cα-C reductive elimination (via TS4), followed by the favorable oxidative addition (via TS1). The process is named as Path-α' (vide infra, Figure 7). To verify the facility of Path-α', the isomerization from Int5 to Int6 was investigated. Taking into account that the interconversion between η3-intermediates is unlikely in light of the configurational reason, we tried to obtain the corresponding η1-intermediates. As shown in Figure 3, the solvent coordinates with Int5 to generate the dioxane-coordinated η3-intermediate Int5s with Cu…dioxane distance of 2.956 Å. Then, in the presence of dioxane, Cu-Cγ bond gradually dissociates and Cu-Cα bond forms to give η1-intermediate α-Int5s. Similarly, the gradual Cu-Cα bond cleavage and Cu-Cγ bond formation gives γ-Int5s. The free energies of them are 11.1 and -6.1 kcal/mol, and the isomerization via γ-Int5s was then investigated since the lower free energy. The strong trans effect of CH2Bpin in α-Int5s weakens the Cu-Cα bond, which contributes to its high free energy.

10


Page 10 of 24

Page 11 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The energy scan of the process from Int5s to γ-Int5s shows the highest energy point of -0.9 kcal/mol.29 γ-Int5s then undergoes barrierless 1,3-conversion to generate the Cα-Cu bonded η1-intermediate Int5s' with decreased free energy of -9.0 kcal/mol.29 This facile conversion might be due to the stability of Int5s', which is derived from the conjugation of benzene ring and double bond. Next, dioxane dissociation from Int5s' automatically generates η3-intermediate Int6 with energy decrease of 8.7 kcal/mol. Therefore, successive dioxane coordination, η1-intermediate formation, 1,3-convertion and dioxane dissociation constitute a feasible isomerization pathway, with highest energy point of -0.9 kcal/mol. Accordingly, Path-α' and Path-γ share the favorable oxidative addition step. After that, isomerization & Cα-C reductive elimination or direct Cγ-C reductive elimination process delivers α or γ-product, with highest energy point of -0.9 kcal/mol and -4.6 kcal/mol respectively (vide infra, Figure 7). Therefore, the reductive elimination determines the relative facility of the above two processes. Path-γ is more favorable than Path-α', with free energy barrier difference of 3.7 kcal/mol. Alkene insertion-β-elimination mechanism

O B O

O O B

Cu

b

O Li

O

O

Cu

O B

EtO

P

O Lib

EtO

TSAdd

O

O B Cu

Cu O

Lia O

O

P

O

Lia

O

OEt

12.6 OEt

-9.3

1a

Int1 IntAdd

-20.9 cat3 O B O

O

Lib Lia O

O

-25.7

Cu O

Cu

O B

O B O

O B

O Lib

Cu

Cu O EtO

P

O

Lia

O

OEt

Figure 4. The energy profile of alkene insertion step Taking the γ-alkylation process for example, we further investigated the alkene insertion-β-elimination mechanism. As shown in Figure 4, we first tried to obtain the 11



alkene insertion transition state TSAdd from intermediate Int1. However, all attempts at locating the transition state failed, and the Cu-CH2Bpin bond always reforms and the C-PO4Et2 bond breaks automatically during the geometry optimization. This observation indicates that the Cu-CH2Bpin bond break is difficult and the -PO4Et2 group tends to leave. To estimate the energy demand of the alkene insertion step, we investigated the free energy of the fixed optimization of TSAdd. The energy is calculated to be 12.6 kcal/mol. Therefore, the energy barrier of alkene insertion is about 33.5 kcal/mol (cat3→TSAdd), which is larger than the overall energy barrier of the overall energy barrier of the oxidative addition-reductive elimination mechanism (21.5 kcal/mol, Path-γ). Accordingly, both geometry optimization and energy estimation indicate that the alkene insertion process is difficult. Therefore, the alkene insertion-β-elimination process is unfavorable compared with oxidative addition-reductive elimination mechanism. In the following sections we only consider the favorable oxidative addition-reductive elimination mechanism. 3.2 Mechanism with solvent DMF (condition 2)

12


Page 12 of 24

Page 13 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5. Energy profiles of reaction mechanism with DMF as solvent Figure 5 shows the energy profiles of Path-α (brown line) and Path-γ (blue line) in solvent DMF. Similar to the mechanism of solvent dioxane, cat3DMF first coordinates with 1aDMF to give Int1DMF and Int2DMF with increased free energies of -9.5 and 5.0 kcal/mol. Then, oxidative addition occurs via TS1DMF and TS2DMF to generate Cu(III) intermediates Int3DMF and Int4DMF, with energy barriers of 22.4 and 37.1 kcal/mol, respectively. Therefore, the former is more favorable, and the CH2Bpin group still occupies the trans position of the leaving group. Then, Int3DMF and Int4DMF dissociate the LiCuLi structure to give Int5DMF and Int6DMF. The Cγ-C and Cα-C reductive elimination subsequently occur from them via TS3 DMF and TS4 DMF with free energies of -0.9 and -10.5 kcal/mol, respectively. After that, the γ- and α-alkylation product 3aDMF and 4aDMF were yielded. With the regeneration of CuOMe, the free energies decrease to -38.0 and -42.2 kcal/mol, respectively. Similar to the mechanism of solvent dioxane, both Path-γ and Path-α are irreversible, and the oxidative addition step determines the relative facility. Since the oxidative addition via TS1DMF is favorable than that via TS2DMF, Path-γ is favorable than Path-α. However, it is not consistent with the α-selectivity for DMF solvent in experiment, which further indicates a more favorable pathway for α-alkylation. Based on the calculation of Section 3.1, we investigated the facility of Path-α' for DMF solvent (successive oxidative addition via TS1DMF, isomerization from Int5 DMF to Int6 DMF, and reductive elimination via TS4 DMF).

13



Page 14 of 24

Figure 6. The isomerization process from Int5DMF to Int6DMF Figure 6 shows the isomerization from Int5sDMF to Int6sDMF. DMF first coordinates with Int5DMF to generate Int5sDMF with Cu…DMF distance of 2.664 Å. In the presence of DMF, the η1-intermediates α-Int5sDMF and γ-Int5sDMF are generated with free energies of 8.6 and -2.5 kcal/mol. Since the former is already higher than that of TS3DMF (-0.9 kcal/mol, Cγ-C reductive elimination transition state), the isomerization via γ-Int5sDMF was investigated. The energy scan of the process from Int5s to γ-Int5s shows the highest energy point of -2.4 kcal/mol.30 γ-Int5sDMF then undergoes facile 1,3-conversion to generate Int5s'

DMF

with decreased free

energy of -8.6 kcal/mol.30 Int5s' then dissociates DMF generating Int6 with energy decrease of 9.5 kcal/mol. Therefore, successive dioxane coordination, η1-intermediate formation, 1,3-convertion and dioxane dissociation constitute a feasible isomerization pathway, with highest energy point of -2.4 kcal/mol. 3.3 Mechanism comparison between dioxane and DMF According to the calculation results in Section 3.1 and 3.2, the favorable mechanisms for γ- and α-alkylation are obtained. Herein, this section shows the overall mechanisms and the mechanism comparison of the dioxane and DMF situations. 14


Page 15 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


O B O

O O B

O Lib

O O B

Cu O

Cu

Lia O EtO

P

O OEt -0.9

TS1 -2.7 -20.9

-4.6 TS3 -7.0

-14.8 Int5

Int1 -9.3

cat3

O Cu

Int6 -17.7

TS4

Int3 -26.1 -29.1 3a

alkene coordination

-32.4 4a

C -C reductive elimination

oxidative addition

isomerization

&


Figure 7. The overall mechanism for solvent dioxane Figure 7 shows the overall mechanisms for solvent dioxane. Path-α' and Path-γ share the favorable oxidative addition step. After that, isomerization & Cα-C reductive elimination or direct Cγ-C reductive elimination process delivers α or γ-product, with highest energy point of -0.9 kcal/mol and -4.6 kcal/mol. Therefore, the reductive elimination determines the relative facility of the above two processes. Path-γ is more favorable than Path-α', which is consistent with the γ-alkylation selectivity in experiment. What’s more important, the smaller energy barrier difference of 3.7 kcal/mol (-0.9 vs -4.6 kcal/mol) is more reasonable than that of 15.5 kcal/mol (vide supra).

15



Page 16 of 24

O B

O

O O B

Cu

O Lib

O

Cu

O

Lia O 1.5 Int1DMF -9.5

EtO

P

TS3DMF -0.9 -14.2 Int5DMF

-1.6 TS4DMF -18.1

-20.9 cat3DMF

O

Cu

O OEt

TS1DMF

B

O

-10.5

Int6DMF -21.9 Int3DMF

3aDMF -38.0 4aDMF -42.2

alkene coordination

oxidative addition

isomerization

&



Figure 8. Favorable mechanism with solvent DMF Figure 8 shows the overall mechanisms for solvent DMF. Similar to the condition of solvent dioxane, Path-γ and Path-α' share the same oxidative addition step. Then, direct Cγ-C reductive elimination and the isomerization & Cα-C reductive elimination occur for Path-γ and Path-α' to generate the γ-alkylation and α-alkylation products, respectively. The highest energy points of the direct Cγ-C reductive elimination and isomerization & Cα-C reductive elimination processes are -0.9 and -1.6 kcal/mol. Therefore, the latter is slightly more feasible than the former, indicating Path-α' (rather than Path-γ) is favorable for solvent DMF. The calculation is consistent with the α-alkylation selectivity in DMF solvent. Overall, the consistency further verifies the validity of Path-α' for α-alkylation process. In summary, for dioxane solvent, direct Cγ-C reductive elimination is feasible than isomerization & Cα-C reductive elimination, which leads to γ-alkylation product. For DMF solvent, isomerization & Cα-C reductive elimination is feasible than direct Cγ-C reductive elimination, which leads to α-alkylation product. Comparison shows that the solvent interferes the isomerization between the two reductive elimination precursors, thereby affects the relative facility of the α- and γ-alkylation. Specifically, the DMF solvent promotes the isomerization (from Int5DMF to Int6DMF) to make the 16


Page 17 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


isomerization & Cα-C reductive elimination process feasible. Meanwhile, dioxane could not promote the isomerization (from Int5 to Int6), and thus the direct Cγ-C reduction elimination from Int5 is kinetically favored. In addition, successive solvent coordination, η1-intermediate formation, 1,3-convertion and solvent dissociation steps constitute a feasible isomerization pathway, among which η1-intermediate formation step is rate-determining. The promotion effect of DMF originates from its stronger electron-donating ability than dioxane to facilitate the η1-intermediate formation (Int5DMF→γ-Int5sDMF). In conclusion, the solvent-controlled α/γ-selectivity is determined by the different electron-donating ability of the solvent. 3.4 Mechanism with dioxane and NHC

O B

NHC

O O

LiCl O O P OEt Int4NHC EtO Cu

B

NHC

LiCl

Int1NHC -6.6

-8.6

Int2NHC -6.7

BpinCH2 Cu NHC

Cl Li

Cu O O P OEt EtO

Cl Cu NHC 0.0

O

7.6 Int4NHC -22.7

O O B

-21.4 TS3NHC Int5NH C -27.5

TS1NHC -25.2 Int3NHC

O O B

NHC Cu

LiCl O

O P Int3NHC EtO

OEt

alkene coordination

O O B

LiPO4Et2

O

Cu

LiCl

Cu

O

O

B

NHC

P OEt

O

Cl Cu NHC

-27.3 TS4 NHC

-29.3 Int6N

NHC

TS1NHC EtO

OEt P O OEt Cu NHC O

TS2NHC -9.0

-14.2

LiCl

Cl Li O OEt P O O OEt O B Cu NHC

3a -58.1

HC

Li

Cl

O O P OEt EtO

oxidative addition

O B

NHC Cu

O

Li

Cl

-61.4 4a

O O P OEt EtO

reductive elimination

Figure 9. Energy profiles of the reaction with dioxane solvent and NHC ligand Similar to the calculation in Section 3.1 and 3.2, the Path-γ and Path-α were also firstly investigated for the NHC-Cu(I) catalyzed reaction in dioxane solvent. As shown in Figure 9, active catalyst NHC-Cu-CH2Bpin was generated from NHC-Cu-Cl with decreased free energy of -8.6 kcal/mol. 28,31 The catalyst then coordinates with 1a to give Int1NHC and Int2NHC with slight energy increase of 2.0 and 1.9 kcal/mol. 17



Inspired by the Li cation assisted -PO4Et2 group leaving in the oxidative addition of Scetion 3.1 and 3.2, we put LiCl around the -PO4Et2 group of Int1NHC and Int2NHC to generate Int3NHC and Int4NHC with decreased free energies of -25.2 and -22.7 kcal/mol. Then, SN2′-type oxidative addition occurs via transition states TS1NHC and TS2NHC with free energies of -14.2 and -9.0 kcal/mol. The former oxidative addition process is always more favorable, and the CH2Bpin group occupies the trans position of the leaving group. After oxidative addition, the Cu(III) intermediates Int5NHC and Int6NHC are generated with decreased free energies of -27.5 and -29.3 kcal/mol. Notably, one oxygen atom of the -PO4Et2 group coordinates tightly with the Cu(III) center in Int5NHC and Int6NHC. The corresponding Cγ-C and Cα-C reductive elimination then occur via transition states TS3NHC and TS4NHC with free energies of -21.4 and -27.3 kcal/mol. They are still lower than that of TS1NHC and TS2NHC (-14.1 and -9.0 kcalmol), indicating the irreversibility of Path-γ and Path-α. Finally, the γ-alkylation product 3a and α-alkylation 4a are yielded with the regeneration of catalyst NHC-Cu-Cl. We also tried to investigate the isomerization of reductive elimination precursors Int5NHC and Int6NHC. However, the coordinated -PO4Et2 group occupies the coordination site and disables the solvent coordination to Cu center. We tried to remove the -PO4Et2 group and create a coordination site for the solvent. Unfortunately, this process requires the high energy demand of 33.9 kcal/mol. It is understandable because the Cu center and PO4Et2 group are respectively positive and negative in the presence of neutral NHC ligand, the separation of them is difficult in low polarity solvent. In addition, the -PO4Et2 group still coordinates with the Cu center in the following reductive elimination. Therefore, with the coordination of neutral NHC ligand, the leaving group tightly coordinates with the Cu center after the oxidation addition, preventing the solvent coordination, and thereby blocking the isomerization of reductive elimination precursors. Namely, Path-α' is impossible with the NHC ligand coordination.

18


Page 18 of 24

Page 19 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Accordingly, Path-γ and Path-α are the favorable mechanism for γ- and α-alkylation. Both Path-γ and Path-α are irreversible, indicating that the oxidative addition step determines the relative facility of them. The energy barrier difference of these two oxidative addition steps is determined by the energy difference between TS1NHC and Int1NHC (7.6 kcal/mol). The energy barrier difference (7.6 kcal/mol) is larger than that of only dioxane situation (3.7 kcal/mol, Section 3.1), which is consistent well with the increased selectivity after the addition of NHC ligand in experiment.

4. Conclusion The Suzuki-Miyaura coupling (SMC) reaction with the novel nucleophile of gem-diboryalkane is a powerful synthetic method to construct C(sp3)-C bonds. Recently, our group accomplished the coupling between gem-diboryalkane and allyl electrophiles by Cu-catalyzed system. Experimentally, γ-alkylation product was obtained in dioxane and the dioxane & NHC ligand situation, while α-alkylation product was obtained in DMF solvent. The DFT calculations were carried out to illustrate the detailed mechanism and origin of the solvent and NHC ligand controlled α/γ-regioselectivity. For both of dioxane and DMF, γ-alkylation undergoes oxidative addition (CH2Bpin trans to leaving group) and direct Cγ-C reductive elimination (Path-γ). The α-alkylation is found to undergo oxidative addition, isomerization and Cα-C reductive elimination (Path-α') rather than oxidative addition (-CH2Bpin cis to the leaving group) and Cα-C reductive elimination (Path-α). Path-γ and Path-α' is respectively favorable for dioxane and DMF solvent, which is consistent with the γand α-selectivity in experiment. The successive solvent coordination, η1-intermediate formation, 1,3-convertion and solvent dissociation constitute a feasible isomerization pathway. The solvent interferes the isomerization step, thereby affects the relative facility of the α- and γ-alkylation processes. The stronger electron-donating ability of DMF than dioxane promote the isomerization to facilitate α-alkylation. In the presence of NHC ligand, Path-γ and Path-α are the favorable mechanism for 19



γ-alkylation and α-alkylation. Path-α' is unfavorable compared with Path-α because the difficult isomerization step, which is derived from the tight coordination of -PO4Et2 group to Cu center. The γ-regioselectivity is determined by the relative facility of the oxidative addition step (in which -CH2Bpin trans to the leaving group). ASSOCIATED CONTENT Supporting Information Details of different forms of Cu(I) catalyst, Complete content for reference (27), (27) and (30), and Cartesian coordinates, free energies, and thermal corrections. The Supporting Information is available free of charge on the ACS Publications website.

AUTHOR INFORMATION Corresponding Author [email protected]; [email protected] Notes The authors declare no competing financial interests.

ACKNOWLEDGMENT We thank the 973 Program (2012CB215305), NSFC (21325208, 21402181, 21572212, 21702041, 51402078), IPDFHCPST (2014FXCX006), CAS (KFJ-EW-STS-051, YZ201563), FRFCU, PCSIRT, and Young Scholar Enhancement Foundation (Plan B) of Hefei University of Technology (JZ2016HGTB0711). The numerical calculations in this paper have been done on the supercomputing system in the Supercomputing Center of University of Science and Technology of China.

REFERENCES (1) For reviews of the Suzuki-Miyaura coupling (SMC) reaction, see: (a) Miyaura, N.; Suzuki, A. Chem. Rev. 1995, 95, 2457. (b) Chemler, S. R.; Trauner, D.; Danishefsky, S. J. Angew. Chem., Int. Ed. 2001, 40, 4544. (c) Lloyd-Williams, P.; Giralt, E. Chem. Soc. Rev. 2001, 30, 145. (2) For reviews of reactions with gem-diborylalkanes, see: Xu, L.; Zhang, S.; Li, P. 20


Page 20 of 24

Page 21 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Chem. Soc. Rev. 2015, 44, 8848. (3) (a) Nakamura, M.; Hara, K.; Hatakeyama, T.; Nakamura, E. Org. Lett. 2001, 3, 3137. (b) Hatakeyama, T.; Nakamura, M.; Nakamura, E. J. Am. Chem. Soc. 2008, 130, 15688. (4) Endo, K.; Ohkubo, T.; Hirokami, M.; Shibata, T. J. Am. Chem. Soc. 2010, 132, 11033. (5) (a) Li, H.; Zhang, Z.; Shangguan, X.; Huang, S.; Chen, J.; Zhang, Y.; Wang, J. Angew. Chem., Int. Ed. 2014, 53, 11921. (b) Lee, J. C.; McDonald, R.; Hall, D. G. Nat. Chem. 2011, 3, 894. (c) Feng, X.; Jeon, H.; Yun, J. Angew. Chem., Int. Ed. 2013, 52, 3989. (6) (a) Endo, K.; Ohkubo, T.; Shibata, T. Org. Lett. 2011, 13, 3368. (b) Endo, K.; Ohkubo, T.; Ishioka, T.; Shibata, T. J. Org. Chem. 2012, 77, 4826. (7) (a) Sun, C.; Potter, B.; Morken, J. P. J. Am. Chem. Soc. 2014, 136, 6534. (b) Potter, B.; Szymaniak, A. A.; Edelstein, E. K.; Morken, J. P. J. Am. Chem. Soc. 2014, 136, 17918. (8) Hong, K.; Liu, X.; Morken, J. P. J. Am. Chem. Soc. 2014, 136, 10581. (9) Zhang, Z.-Q.; Yang, C.-T.; Liang, L.-J.; Xiao, B.; Lu, X.; Liu, J.-H.; Sun, Y.-Y.; Marder, T. B.; Fu, Y. Org. Lett. 2014, 16, 6342. (10) Kim, J.; Park, S.; Park, J.; Cho, S. H. Angew. Chem., Int. Ed. 2016, 55, 1498. (11) Zhang, Z.-Q.; Zhang, B.; Lu, X.; Liu, J.-H.; Lu, X.-Y.; Xiao, B.; Fu, Y. Org. Lett. 2016, 18, 952. (12) (a) Yoshikai, N.; Zhang, S.-L.; Nakamura, E. J. Am. Chem. Soc. 2008, 130, 12862. (b) Yamanaka, M.; Kato, S.; Nakamura, E. J. Am. Chem. Soc. 2004, 126, 6287. (13) Karlström, A. S. E.; Bäckvall, J.-E. Chem.-Eur. J. 2001, 7, 1981 (14) Bartholomew, E. R.; Bertz, S. H.; Cope, S.; Murphy, M.; Ogle, C. A. J. Am. Chem. Soc. 2008, 130, 11244. (15) (a) Nagao, K.; Yokobori, U.; Makida, Y.; Ohmiya, H.; Sawamura, M. J. Am. Chem. Soc. 2012, 134, 8982. (b) Ohmiya, H.; Yokobori, U.; Makida, Y.; Sawamura, M. J. Am. Chem. Soc. 2010, 132, 2895. (c) Hojoh, K.; Shido, Y.; Ohmiya, H.; Sawamura, M. Angew. Chem. Int. Ed. 2014, 53, 4954. 21



(16) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Keith, T.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, revision D.01; Gaussian, Inc.: Wallingford, CT, 2013. (17) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (b) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B: Condens. Matter Mater. Phys. 1988, 37, 785. (18) B3LYP has been frequently used in previous theoretical studies for transition metal-catalyzed systems; see: (a) Bagchi, V.; Paraskevopoulou, P.; Das, P.; Chi, L.; Wang, Q.; Choudhury, A.; Mathieson, J. S.; Cronin, L.; Pardue, D. B.; Cundari, T. R.; Mitrikas, G.; Sanakis, Y.; Stavropoulos, P. J. Am. Chem. Soc. 2014, 136, 11362. (b) Bukharov, M. S.; Shtyrlin, V. G.; Mamin, G. V.; Stapf, S.; Mattea, C.; Mukhtarov, A. S.; Serov, N. Y.; Gilyazetdinov, E. M. Inorg. Chem. 2015, 54, 9777. (c) Markovic, M.; Ramek, M.; Loher, C.; Sabolovic, J. Inorg. Chem. 2016, 55, 7694. (d) Xu, Z.-Y.; Yu, H.-Z.; Fu, Y. Chem. Asian. J. 2017, 12, 1765. (e) Zhang, Q.; Zhang, Z.-Q.; Fu, Y.; Yu, H.-Z. ACS Catal. 2016, 6, 798. (f) Zhang, Q.; Yu, H.-Z.; Li, Y.-T.; Liu, L.; Huang, Y.; Fu, Y. Dalton Trans. 2013, 42, 4175. (19) Hollwarth, A.; Bohme, M.; Dapprich, S.; Ehlers, A. W.; Gobbi, A.; Jonas, V.; Kohler, K. F.; Stegmann, R.; Veldkamp, A.; Frenking, G. Chem. Phys. Lett. 1993, 208, 237. (20) Gonzalez, C.; Schlegel, H. B. J. Phys. Chem. 1990, 94, 5523. 22


Page 22 of 24

Page 23 of 24 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(21) (a) Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215. (b) Zhao, Y.; Truhlar, D. G. Acc. Chem. Res. 2008, 41, 157. (22) Ehlers, A. W.; Bohme, M.; Dapprich, S.; Gobbi, A.; Hollwarth, A.; Jonas, V.; Kohler, K. F.; Stegmann, R.; Veldkamp, A.; Frenking, G. Chem. Phys. Lett. 1993, 208, 111. (23) Hollwarth, A.; Bohme, M.; Dapprich, S.; Ehlers, A. W.; Gobbi, A.; Jonas,V.; Kohler, K. F.; Stegmann, R.; Veldkamp, A.; Frenking, G. Chem. Phys. Lett. 1993, 208, 237. (24) The M06//B3LYP method has been frequently used in transition-metal theoretical studies; for examples, see: (a) Lin, M.; Kang, G.-Y.; Guo,Y.-A.; Yu, Z.-X. J. Am. Chem. Soc. 2012, 134, 398. (b) Herbert, M. B.; Lan, Y.; Keitz, B. K.; Liu, P.; Endo, K.; Day, M. W.; Houk, K. N.; Grubbs, R. H. J. Am. Chem. Soc. 2012, 134, 7861. (c) Hong, X.; Stevens, M. C.; Liu, P.; Wender, P. A.; Houk, K. N. J. Am. Chem. Soc. 2014, 136, 17273. (d) Zhang, Q.; Yu, H.-Z.; Fu, Y. Organometallics 2013, 32, 4165. (e) Jiang, Y.-Y.; Yu, H.-Z.; Fu, Y. Organometallics 2013, 32, 926. (f) Li, Y.; Liu, S.; Qi, Z.; Qi, X.; Li, X.; Lan, Y. Chem. - Eur. J. 2015, 21,10131.(g) Qi, X.; Zhang, H.; Shao, A.; Zhu, L.; Xu, T.; Gao, M.; Liu, C.; Lan, Y. ACS Catal. 2015, 5, 6640. (25) Yoshikai, N.; Nakamura, E. Chem. Rev. 2012, 112, 2339. (26) See Supporting Information for more details of different forms of Cu(I) catalyst. (27) The corresponding oxidative addition transition states TS1' and TS2' without the bridging structure Lia…BpinCH2CuOMe…Lib were also located to investigated the roles of the bridging structure. See Supporting Information for more details. (28) (a) Ohishi, T.; Nishiura, M.; Hou, Z. Angew. Chem., Int. Ed. 2008, 47, 5792. (b) Ohmiya, H.; Yokokawa, N.; Sawamura, M. Org. Lett. 2010, 12, 2438. (c) Whittaker, A. M.; Rucker, R. P.; Lalic, G. Org. Lett. 2010, 12, 3216. (d) Shintani, R.; Takatsu, K.; Takeda, M.; Hayashi, T. Angew. Chem., Int. Ed. 2011, 50, 8656. (d) Takeda, M.; Takatsu, K.; Shintani, R.; Hayashi, T. J. Org. Chem. 2014, 79, 2354. (29) See Supporting information for the energy scan of the transformation process from Int5s to γ-Int5s, and the process from γ-Int5s to Int5s'. (30) See Supporting information for the energy scan of the transformation process 23



from Int5sDMF to γ-Int5sDMF, and the process from γ-Int5sDMF to Int5s'DMF. (31) Kleeberg, C.; Dang, L.; Lin, Z.; Marder, T. B. Angew. Chem., Int. Ed. 2009, 48, 5350.

24


Page 24 of 24

Mechanistic Insights into Solvent and Ligand Dependency in Cu(I

Recommend Documents