UvA-DARE is a service provided by the library of the University of Amsterdam (http://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

Importance of the Reducing Agent in Direct Reductive Heck Reactions

Raoufmoghaddam, S.; Mannathan, S.; Minnaard, A.J.; de Vries, J.G.; de Bruin, B.; Reek,J.N.H.Published in:ChemCatChem

DOI:10.1002/cctc.201701206

Link to publication

Citation for published version (APA):Raoufmoghaddam, S., Mannathan, S., Minnaard, A. J., de Vries, J. G., de Bruin, B., & Reek, J. N. H. (2018).Importance of the Reducing Agent in Direct Reductive Heck Reactions. ChemCatChem, 10(1), 266-272.https://doi.org/10.1002/cctc.201701206

General rightsIt is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s),other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, statingyour reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Askthe Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam,The Netherlands. You will be contacted as soon as possible.

Download date: 29 Jan 2021

Importance of the Reducing Agent in Direct ReductiveHeck ReactionsSaeed Raoufmoghaddam,[a] Subramaniyan Mannathan,[b] Adriaan J. Minnaard,[b]

Johannes G. de Vries,[b, c] Bas de Bruin,[a] and Joost N. H. Reek*[a]

Introduction

Despite the massive progress in Heck coupling reactions overthe last decades,[1, 2] the thorough mechanistic understanding

of the “direct” catalytic reductive Heck reaction remains ellu-sive, as it is a rather unexplored reaction. Conversely, the uti-lization of highly reactive organometallic reagents to obtain re-lated products has been studied extensively.[3, 4] Although nu-

merous advances have been made in Pd-catalyzed conjugateaddition with organometallic reagents (Grignards, organozincs,

and boronic acids) in the past decade,[5–7] the direct Pd-cata-lyzed reductive Heck reaction of alkenes with aryl halides has

hardly been investigated.[8–12] In general, the replacement ofexpensive air- and moisture- sensitive organometallic reagentswith aryl halides is very tempting for industrial synthetic trans-formations. The reductive arylation product in the Pd-catalyzed

reaction was first reported by Cacchi, and it was reasoned thatthe presence of trialkylamines and formic acid facilitate the re-

duction step by acting as the hydride source to form the hydri-

do-palladium-alkyl complex.[8, 9] The reductive Heck product isthen produced through the reductive elimination of the com-

plex, whereas b-hydride elimination of the palladium-alkylcomplex forms the Mizoroki–Heck product (Scheme 1).

It was also proposed that formic acid plays two importantroles: protonation of the Pd-alkyl species and reduction of PdII

to Pd0. Alternatively, we reported the reductive Heck reaction

using an N-heterocyclic carbene (NHC) palladium complex, tri-alkylamines (i.e. , N,N-diisopropylethylamine; DIPEA), and aprot-

ic solvents (i.e. , N-methyl-2-pyrrolidone; NMP).[10, 12]

The nature of the base and solvent play a crucial role in the

reaction and both affect the catalytic results by changing theregioselectivity of the reaction. Additionally, we found that the

The role of the reductant in the palladium N-heterocyclic car-bene (NHC) catalyzed reductive Heck reaction and its effect on

the mechanism of the reaction is reported. For the first time inthis type of transformation, the palladium-NHC-catalyzed re-ductive Heck reaction was shown to proceed in the presenceof LiOMe and iPrOH even at 10 8C to give the products very ef-ficiently in excellent yields and with exceptional chemoselectiv-ities. This study shows that the reaction proceeds through two

distinct mechanisms that depend on the nature of the reduc-ing agent. In the presence of a protic solvent or acidic mediumthe reaction undergoes protonation to yield the reduced prod-uct, whereas in the absence of proton source, it proceedsthrough the insertion of the reductant followed by reductive

elimination. The kinetic data reveal that the oxidative additionis the rate-determining step in the reaction. The reaction pro-

files show first-order kinetics in aryl iodide and Pd and zero-order kinetics in LiOMe, benzylideneacetone, and the excess

amount of NHC ligand. In addition, the reaction progress kinet-

ic analysis shows that neither catalyst decomposition nor prod-uct inhibition occurs during the reaction. DFT calculations of

the key steps confirm that the oxidative addition step is the

rate-determining step in the reaction. Deuterium-labeling ex-

periments indicate that the product is formed by the protona-tion of the Pd@Calkyl bond of the intermediate formed afterenone insertion into the Pd@CAr bond. Application of chiralNHC ligands in the asymmetric reductive Heck reaction onlyresults in poor enantioselectivities (enantiomeric excess up to20 %) and is also substrate specific. DFT calculations suggest

that the migration of the aryl group to the alkene of the sub-strate is the enantioselectivity-determining step of the reac-tion. It is further shown that if the steric bulk at the enone issmall (a methyl group), the two transition state barriers from[PdII(L2)(ArI)(enone)] species Cre and Csi, which have the re and

si face of the enone substrate coordinated to Pd, are very simi-lar, in line with the experimental results. With a slightly larger

group (an isopropyl substituent) a significant difference inenergy barriers is calculated (2.6 kcal mol@1), and in the experi-

ment this product is formed with a modest enantiomeric

excess (up to 20 %).

[a] Dr. S. Raoufmoghaddam, Prof. B. de Bruin, Prof. J. N. H. ReekVan’t Hoff Institute for Molecular SciencesUniversity of AmsterdamScience Park 904, 1098 XH, Amsterdam (The Netherlands)E-mail : J.N.H.Reek@uva.nl

[b] Dr. S. Mannathan, Prof. A. J. Minnaard, Prof. J. G. de VriesStratingh Institute for ChemistryUniversity of GroningenNijenborgh 7, 9747 AG, Groningen (The Netherlands)

[c] Prof. J. G. de VriesLeibniz-Institut fer Katalyse e.V. an der Universitat RostockAlbert-Einstein-Strasse 29a, 18059 Rostock (Germany)

Supporting information and the ORCID identification number(s) for theauthor(s) of this article can be found under https://doi.org/10.1002/cctc.201701206.

ChemCatChem 2018, 10, 266 – 272 T 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim266

Full PapersDOI: 10.1002/cctc.201701206

presence of DIPEA as both a reductant and base helps the re-ductive cleavage step and, consequently, promotes the forma-

tion of the reductive Heck product. Based on our observationsin the previous studies, we proposed that the reaction does

not involve the protonation of the palladium-alkyl species toyield the reductive Heck product. Instead, the role of DIPEA in

the reaction is to act as a reductant through its coordination

to the palladium-alkyl species (after iodide dissociation) fol-lowed by b-hydride elimination of the DIPEA to form the hy-

dride species. The reductive Heck product is formed in thefinal reductive elimination step. This role of DIPEA was con-

firmed by performing an experiment using deuterium-labeledtrialkylamine ([D15]Et3N), which resulted in the formation of the

deuterated reductive Heck product.[12] This indicates that the

hydride source depends on the nature of the solvent (protic oraprotic) and base (reductant or nonreductant) applied in such

a reaction.[13–15] In addition, studies on the asymmetric reduc-tive Heck reaction[15, 16] have been reported by Buchwald et al.

who used aryl triflates in an intramolecular fashion to synthe-size chiral 3-substituted indanones with moderate enantiomer-ic excess (ee) values in most cases.[17] More recently, an enantio-

selective palladium-phosphine-catalyzed intramolecular reduc-tive Heck reaction of aryl halides was reported to afford 3-ary-lindanones with moderate to high ee values.[18] It was proposedthat the addition of trialkylammonium salts in ethylene glycol

facilitates the halide dissociation through its hydrogen bonddonor capacity. Another enantioselective transformation of this

class was reported with the asymmetric arylative dearomatiza-tion of indoles through palladium-phosphine-catalyzed reduc-tive Heck reaction using sodium formate in methanol.[19] To

date, the asymmetric reductive Heck reaction is rather limitedto intramolecular reactions, which is particularly interesting for

the synthesis of natural products.[17–24] To the best of ourknowledge, there are no examples that describe the “direct in-

termolecular” enantioselective reductive Heck reaction with a

high ee, most likely because of the challenges involved in theasymmetric induction step.

In this contribution, we report the palladium-NHC-catalyzeddirect reductive Heck reaction of para-substituted benzylide-

neacetones with 4-iodoanisole using isopropanol as both thesolvent and reductant. The reaction proceeds very efficiently

even at low temperatures (10 8C). Additionally, DFT calculations,kinetics, and other mechanistic studies were performed to pro-

vide a detailed mechanistic picture of the reaction. We alsoprobed the direct intermolecular asymmetric reductive Heck

reaction and elucidated the enantioselectivity-determiningstep by performing DFT calculations.

Results and Discussion

Initial optimizations

To explore the role of the reductant we studied the typical re-

ductive Heck reaction between benzylideneacetone (1) and 4-iodoanisole (2) that resulted in the formation of the expected

product 3 with traces of the Heck product 4 (obtained as amixture of E/Z isomers) and 4,4’-dimethoxybiphenyl (5 ;

Table 1). We used our reaction conditions optimized previous-

ly[12] as a starting point to explore reaction conditions basedon various solvents in different combinations with various

bases. These experiments revealed that isopropanol as a sol-

Scheme 1. Plausible mechanistic pathways ; Heck product (by b-hydrideelimination) versus reductive Heck product (by protonation or reductiveelimination of [PdII(alkyl)(H)] species).

Table 1. Pd-catalyzed reductive arylation.[a]

Entry Base t Solvent Conv.[b] Yield [%][b]

[equiv.] [h] [v/v] [%] 3 4 5

1 DIPEA (2) 4 NMP 100 90 8 12 K2CO3 (1) 20 NMP 96 0 96 03 K2CO3 (1) 20 NMP/iPrOH (1.5:0.5) 87 13 74 04 K2CO3 (1) 20 NMP/iPrOH (1:1) 75 32 43 05 K2CO3 (1) 20 NMP/iPrOH (0.5:1.5) 66 48 18 06 K2CO3 (1) 20 NMP/iPrOH (0.25:1.75) 64 50 14 07 K2CO3 (1) 20 NMP/iPrOH (0.1:1.9) 63 53 10 08 K2CO3 (1) 20 acetone/iPrOH (0.25:1.75) 60 54 6 09 K2CO3 (1) 20 H2O/iPrOH (0.25:1.75) 20 12 7 010 K2CO3 (1) 20 iPrOH 65 59 5 011 Cs2CO3 (1) 6 THF/iPrOH (0.25:1.75) 89 66 4 1912[c] LiOMe (1) 0.5 iPrOH 99 93 3 313 LiOMe (1) 1 MeOH/iPrOH (1:1) 100 93 4 314 LiOMe (1) 0.5 EtOH 100 44 4 5215 LiOMe (1) 2 HFIPA 5 4 1 016 LiOMe (1) 2 iBuOH 87 73 13 017 LiOMe (1) 1 PhCH(OH)CH3 100 83 16 018 LiOMe (1) 1 cyclopentanol 81 65 3 1219 – 2 iPrOH 1 – – –20 Na2CO3 (1) 20 iPrOH 60 55 5 021 Cs2CO3 (1) 2 iPrOH 99 82 0 1622 DIPEA (1) 20 iPrOH 16 13 3 023 Et3N (1) 20 iPrOH 13 9 4 024 CsPiv (1) 20 iPrOH 44 3 41 0

[a] Reaction conditions: 1 (1.65 mmol), 2 (1.10 mmol), Pd(L1)(MA)2

(0.75 mol %) (MA = maleic anhydride; L1 = 1,3-bis(2,4,6-trimethylphenyl)i-midazolium), base (1–2 equiv.), T = 60 8C, solvent (1.5 mL), internal stan-dard = decane, t = 0.5–20 h. [b] GC yields and conversions based on 2.[c] Approximately 88 % isolated yield.

ChemCatChem 2018, 10, 266 – 272 www.chemcatchem.org T 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim267

Full Papers

vent in combination with K2CO3 as the base gave a good selec-tivity for the formation of 3 with a moderate conversion (65 %)

after 20 h (Table 1, Entries 1–10). If we changed the base toLiOMe (1 equivalent) it resulted in an excellent yield of 3 (93 %)

after just 30 min (Entry 12). Moreover, if we followed the prod-uct formation over time, we observed that the reaction pro-

ceeds to full conversion within just 30 min with nearly exclu-sive formation of 3. The use of other bases did not further im-

prove the reaction (Entries 20–24), and also the replacement of

isopropanol with other alcohols decreased the activity and se-lectivity (Entries 12–18).

Notably, if we used 1,1,1,3,3,3-hexafluoroisopropyl alcohol(HFIPA) instead of iPrOH, the rate of the reaction decreased

substantially, and almost no substrate conversion was ob-served after 2 h (Entry 15). This can be explained as iPrOH is astronger reductant. The reaction in the absence of base result-

ed in the complete recovery of the starting materials(Entry 19).

The essential role of iPrOH as the solvent/reductant in thereaction was then probed by performing two independent la-beling experiments using [D1]iPrOD and [D7]iPrOH (Scheme 2).The reaction in [D1]iPrOD results in the formation of the mono-

deuterated product, whereas in [D7]iPrOH only the normal

product was detected, which suggests that the product islikely formed by protonation of the palladium-alkyl intermedi-

ate. These results indicate that the reaction proceeds throughthe protonation of the palladium-alkyl complex followed by re-

duction of PdII to Pd0, which is in contrast to that found for re-actions performed in the presence of DIPEA.[12]

A reductive pathway, with proton exchange between the al-

cohol and the product under the applied reaction conditions,is less likely as this proton exchange would also lead to the in-

corporation of two D atoms next to the carbonyl moiety,which is not observed. We next studied the kinetics of the re-

action by monitoring the product formation over time and weprocessed the data using reaction progress kinetic analysis[25]

to provide a typical rate versus substrate concentration plot.The kinetic data revealed first-order kinetics in the palladium-

NHC catalyst and zero-order kinetics in base (LiOMe) and inthe excess of the NHC ligand. Additionally, the reaction prog-

ress kinetic analysis showed that neither catalyst decomposi-tion nor product inhibition plays a significant role during the

reaction under the applied reaction conditions in the timestudied.

DFT calculations

To gain additional insight into the mechanism, we performedDFT calculations to investigate the various pathways of the cat-

alytic cycle (Figure 1). The geometry optimizations were per-formed by using the TURBOMOLE program package[26] using

BP86 functional[27, 28] together with the def2-TZVP basis set.[29]

The given minima comprise no imaginary frequencies, and the

transition states contain only one imaginary negative frequen-cy (for detailed data and all minima and optimized geometries

see the Supporting Information).

From the experimental results, we considered a pathwaythat involves the oxidative addition of the ArI substrate 2, co-

ordination and insertion of the enone substrate 1 into the Pd@CAr bond, followed by protonation of the resulting Pd@Calkyl

bond by a coordinated iPrOH moiety (Figure 1). The energy ofthe various intermediates and transition states indicates that

the oxidative addition is irreversible and the rate-determining

step of the catalytic reaction. This is in agreement with our ki-netic experiments that show first-order kinetics in [2] and zero-

order kinetics in [1] (see Supporting Information for details). Aswe reported previously,[12] the oxidative addition step may pro-

ceed via various transition state (TS) species with slightly differ-ent activation barriers ([TS2]* = [TS1]*++2.9 kcal mol@1), which

depends on the reaction conditions. In this respect, the transi-

tion state of the oxidative addition step can be decreased inenergy (&3 kcal mol@1) by the de-coordination of 1.

In general, the oxidative addition energy profile indicatesthat it is an overall exergonic process to form species B or C.The formation of species C from species B is an endergonicprocess. Species C undergoes migratory insertion to give spe-

cies D, which proceeds via a low-barrier-energy transition state(TS3). Here, a palladium-O-enolate species (D’) can also be

formed through isomerization, but it is energetically less fa-vored (&2.7 kcal mol@1 higher in energy; DG0

298 K (D’) =

@11.9 kcal mol@1). Species D and E can be stabilized by the co-

ordination of iPrOH (&3 kcal mol@1 lower in energy). The nextstep considered in the reaction path is palladium-alkyl proto-

nation to form the product. This is followed by b-hydrogenelimination from alkoxide E to form acetone and the palladi-

um-hydride species F. The computed barrier for this step is low

(12 kcal mol@1 from D (+ iPrOH), which is in good agreementwith the labeling experiments (Scheme 2) and suggests that

protonation is a key step in product formation. Reformation ofthe Pd0 species A is proposed to involve the base-assisted re-

ductive elimination of HI (LiOMe as a base), which is overallslightly endergonic.

Scheme 2. Deuterium-labeling experiment using [D1]iPrOD or [D7]iPrOH inthe reductive Heck coupling reaction. The reaction with [D1]iPrOD only re-sulted in the incorporation of one D in the product, which excludes the pos-sibility of H/D exchange between the product and [D1]iPrOD during the re-action.

ChemCatChem 2018, 10, 266 – 272 www.chemcatchem.org T 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim268

Full Papers

Mechanism of the reaction

From the kinetic data and DFT calculations, we propose the

mechanism depicted in Scheme 3. Several species, possibly in

equilibrium, may be present for the oxidative addition stepwith different energy barriers. We envision that the catalyticcycle starts with the formation of species A and A’. The oxida-tive addition of 2 to palladium(0)-NHC then takes place to give

both species A and A’, and therefore, afford species B or C. Ac-cording to our kinetic data and DFT calculations, the oxidative

addition is the rate-determining step.As reported previously for similar reactions,[30–35] activated al-

kenes slow down the oxidative addition step by coordinating

to the Pd0 species, thereby stabilizing the Pd species throughthe delocalization of the electron density from the electron-

rich Pd0. The resulting species C undergoes migratory insertioninto benzylideneacetone to form species D, which is in equilib-

rium with the energetically less favored palladium-O-enolate

species D’ (&2.7 kcal mol@1 higher in energy). Although thisspecies is less favored in terms of stability, we cannot rule out

that the product is formed by the protonation of this species,which should have a low barrier as it is after the rate- and se-

lectivity-determining step. As we were unable to find the tran-sition state for this step and it was reported that reductive

elimination reactions proceed through Pd@C species,[36] the re-action most likely proceeds via D.

Alternatively, species C could form from a halogen-bridged

dimeric species (analogous to B), in particular, if bulky ligandsare involved, which could lead to the isomeric form of C.[37, 38]

This alternative Pd species C and its transition states with thearyl group positioned cis to the NHC ligand were also calculat-

ed by using DFT (Figure S31) and were demonstrated to besimilar in energy (&0.3 kcal lower in energy). Additionally, the

energy barriers for the migratory insertion step from these spe-cies are similar. As observed in the labeling experiment, proto-nation of the palladium-alkyl complex in iPrOH occurs to form

product 3 and species E. If species D undergoes b-hydrideelimination, the Heck side-product 4 is formed. Species E con-

verts to species F through b-hydride elimination to result inthe formation of acetone, which was also detected experimen-

tally by using GC and GC–MS. The final steps in the mechanism

that ultimately leads to reformation of species A involve b-hy-dride elimination from the alcoholate E to produce acetone (as

observed in the reaction mixture), followed by reductive de-protonation by LiOMe as a base, to form MeOH and LiI. The

abstraction of a proton from a palladium-hydride species by abase can occur without an activation barrier.[39]

Figure 1. Gibbs free energy diagram [DG0298K ; kcal mol@1] of the DFT-computed reductive Heck reaction (BP86, def2-TZVP).

ChemCatChem 2018, 10, 266 – 272 www.chemcatchem.org T 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim269

Full Papers

Asymmetric reductive Heck reactions

With a clear picture of the kinetics and mechanism of the reac-

tion in hand, we next conducted experiments to study thechallenging direct intermolecular asymmetric reductive Heckreaction using chiral NHC ligands (Table 2). Recently, we

showed that this reaction lends itself to an enantioselective in-tramolecular Heck reaction using monodentate phosphorami-dite ligands.[40] The reactions with the model enone 1 (R = Me)using chiral ligand L4 under different conditions (in iPrOH at60 8C and room temperature or in (S)-(@)-1-phenylethanol) re-sulted in <1 % ee (entries 2–4). The mechanism dictates that

the insertion of enone 1 into the Pd@CAr bond of species C isthe asymmetric induction step, which should be controlled bythe chiral NHC ligands. It is anticipated that more sterically hin-

dered enone substrates will be more prone to the induction ofenantioselectivity by the ligand. As such, we explored the reac-

tion by using other enones with bulkier R groups (phenyl, 9-anthracenyl, ethyl, isopropyl, tert-butyl, and 1-adamantyl). In-

terestingly, if we moved from a methyl to an isopropyl group

in the enone substrate, the ee value increased from <1 to 8 %(Entry 8). Unfortunately, the experiments with the bulkier 1-

adamantyl- or tert-butyl-substituted enones with greater stericinfluences gave no product simply because these substrates

are too bulky to coordinate to the Pd center. Consequently,the reaction resulted in the formation of the homocoupling

product (Entries 9–10). Additionally, the reaction with the iPr-substituted enone at lower temperatures (room temperature

and at 10 8C) gave excellent yields of 3 (90–94 %) with slightlyhigher ee values of 16 and 19 % (Entries 11–12). A further de-

crease of the temperature to 0 8C did not result in conversion

as the substrates were recovered unchanged (Entry 13), whichindicates that 10 8C is the lowest reaction temperature at

which the reaction still progresses. The reactions with variouschiral NHC ligands were, therefore, performed at 10 8C, howev-

er, this did not further improve the ee values (Entries 14–20).According to the mechanism supported by our DFT calcula-

tions, the enantioselectivity is determined by an aryl migration

step. The DFT calculations show that the energy difference be-tween the species Cre and Csi, in which the enone is coordinat-

ed to the Pd center with its re and si face, respectively, is verysmall. We next optimized the geometries of species C using

enones with different R groups (methyl and isopropyl) coordi-nated to the Pd center.

Scheme 3. Proposed mechanism of the reductive Heck reaction; empty co-ordination sites may be occupied by solvent molecules or halides.

Table 2. Pd-catalyzed enantioselective reductive arylation.[a]

Entry R Ligand t T Conv.[b] Yield of ee[c]

[h] [8C] [%] 3[b] [%] [%]

1 methyl L1 0.5 60 99 93 02 methyl L4 1 60 78 66 <13[d] methyl L4 4 60 80 63 <14 methyl L4 8 RT 51 50 <15 phenylphenyl L4 1 60 98 91 <16 9-anthracenyl L4 – – – – –7 ethylethyl L4 – – – – –8 Isopropyl L4 2 60 99 96 89[e] tert-butyl L4 20 60 93 0 –10[e] 1-adamantyl L4 20 60 91 0 –11 isopropyl L4 15 RT 96 94 1612 isopropyl L4 20 10 91 90 1913 isopropyl L4 20 0 <2 0 –14 isopropyl L2 20 10 88 82 1415 isopropyl L3 20 10 89 83 1416 isopropyl L5 20 10 90 81 1917 isopropyl L6 20 10 68 65 1718 isopropyl L7 20 60 <1 0 –19 isopropyl L8 20 10 14 13 1720 isopropyl L8 20 RT 49 47 15

[a] Reaction conditions: 1 (1.65 mmol), 2 (1.10 mmol), Pd(L)(MA)2

(0.75 mol %) (MA = maleic anhydride), LiOMe (1 equiv.), T = 0–60 8C, sol-vent = iPrOH (2 mL), t = 0.5–20 h, internal standard = decane. [b] The con-version and yield were determined by using GC based on 2. [c] ee deter-mined by using chiral HPLC. [d] (S)-(@)-1-Phenylethanol used as a solvent.[e] homocoupling product 5 was formed as the major product and otherbyproducts formed.

ChemCatChem 2018, 10, 266 – 272 www.chemcatchem.org T 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim270

Full Papers

To further study the asymmetric induction step of the reac-tion mechanism, we also probed the migratory insertion step

of these two enones into the Pd center through DFT calcula-tions of the corresponding transition states (Figure 2). They

both undergo migratory insertion through exergonic processes

to give species D*. However, the energy barrier for the twotransition states if we start from the species Cre(Me) and Csi(Me)

revealed almost no differences (0.1 kcal mol@1), whereas thetransition states of the species Cre(iPr) and Csi(iPr) had larger

differences of 2.6 kcal mol@1, which indicates that one pathwayis more preferred than the other. This energy barrier difference

observed for the more bulky substrate is in line with the ex-

periment as for this substrate the product was produced in19 % ee.

Notably, the relative energy barriers at ambient or elevatedtemperatures (or because of the absence of explicit solvent ef-

fects) may become even closer, which rationalizes the fact thatwe did not obtain any ee values higher than 20 %.

Conclusions

The effect of the reducing agent on the palladium N-heterocy-clic carbene catalyzed direct reductive Heck reaction of para-

substituted benzylideneacetones with 4-iodoanisole has beeninvestigated. Compared to that of previous studies, the reac-

tion conditions lead to faster reactions as full conversion wasachieved within 30 min at 60 8C. We further showed that thereaction proceeds even at 10 8C. Importantly, the mechanism

of the product-forming step can be different and depends onthe nature of the reducing agent. In the presence iPrOH, the

product is formed through the protonation of the palladium-alkyl intermediate, which is clear from deuterium-labeling ex-

periments. This is in contrast to the reaction in the presence of

N,N-diisopropylethylamine, for which we found that the prod-uct is formed by reductive elimination from the alkyl-hydride

species. The kinetic data indicate that under both conditionsthe oxidative addition is the rate-determining step, which is in

line with the results of our DFT calculations. Various chiral NHCligands were synthesized and evaluated in the asymmetric re-

ductive Heck reaction, which showed no enantiomeric excessfor the benchmark substrate. In line with this, the difference in

the transition state barriers of the crucial aryl migration step ofthe two [PdII(L2)(ArI)(enone)] species Cre and Csi is negligible. In

contrast, the same transition states for the complexes with a

slightly more bulky substrate (isopropyl enone) are 2.6 kcalmol@1 different in energy, and indeed for this reaction signifi-

cant but low enantiomeric excess values (up to &20 %) wereobtained. Further optimization of asymmetric reductive reac-

tion could proceed along these lines.

Experimental Section

General procedure for the reductive arylation reaction.

To a flame-dried Schlenk tube equipped with a stopper and a stir-ring bar, LiOMe (1 m in THF, 1.1 mmol, 1 equiv.) was added. The sol-vent (THF) was then removed under reduced pressure to obtain awhite powder. The Schlenk tube was then charged withPd0(L)(MA)2 (0.75 mol %), 2 (1.1 mmol), and 1 (1.65 mmol) andflushed three times with vacuum and Ar. iPrOH (2 mL) that con-tained decane as an internal standard was then transferred intothe Schlenk tube, which was then placed into a preheated oil bathat 60 8C. Upon reaction completion (after 30 min; judged by usingGC), the reaction mixture was cooled to RT, and a sample of 10 mLwas taken and diluted to 1 mL with dichloromethane and subse-quently analyzed by using GC/GC–MS (for more detailed data, seeSupporting Information).

DFT calculations

All calculations and geometry optimizations were performed byusing the Turbomole program package[26] coupled to the PQSBaker optimizer[41] with the BOpt package[42] at the spin-unrestrict-ed ri-DFT level using the BP86 functional[27, 28] and the def2-TZVPbasis set[29] for the geometry optimizations. All geometries ofminima with no imaginary frequencies and transition states withone imaginary frequency were characterized by calculating theHessian matrix numerically (for detailed information see the Sup-porting Information).

Figure 2. Migratory insertion step and the corresponding transition states of the Pd species (DFT calculations, BP86/def2-TZVP) in the reductive Heck reaction[DG0

298K ; kcal mol@1] . Alternative transition states with the aryl group positioned cis to the NHC ligand were also calculated (Figure S31), which resulted in2.1 kcal energy difference between the species Cre(iPr) and Csi(iPr) and indicates a similar energy barrier (only 0.5 kcal lower than the trans-positioned species)in the migratory insertion step.

ChemCatChem 2018, 10, 266 – 272 www.chemcatchem.org T 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim271

Full Papers

Acknowledgements

This research has been performed within the framework of the

CatchBio Program. The authors gratefully acknowledge the sup-port of the Smart Mix Program of the Netherlands Ministry of

Economic Affairs and the Netherlands Ministry of Education, Cul-

ture and Science, Royal DSM NV, Organon, and Avantium.

Conflict of interest

The authors declare no conflict of interest.

Keywords: carbene ligands · density functional calculations ·palladium · reaction mechanisms · reduction

[1] I. P. Beletskaya, A. V. Cheprakov, Chem. Rev. 2000, 100, 3009.[2] J. P. Corbet, G. Mignani, Chem. Rev. 2006, 106, 2651.[3] D. G. Hall, Boronic Acids: Preparation and Applications in Organic Synthe-

sis Medicine and Materials, Wiley-VCH, 2011.[4] K. Kikushima, J. C. Holder, M. Gatti, B. M. Stoltz, J. Am. Chem. Soc. 2011,

133, 6902.[5] F. Gini, B. Hessen, A. J. Minnaard, Org. Lett. 2005, 7, 5309.[6] F. Gini, B. Hessen, B. L. Feringa, A. J. Minnaard, Chem. Commun. 2007,

710.[7] C. Johansson Seechurn, M. O. Kitching, T. J. Colacot, V. Snieckus, Angew.

Chem. Int. Ed. 2012, 51, 5062; Angew. Chem. 2012, 124, 5150.[8] S. Cacchi, A. Arcadi, J. Org. Chem. 1983, 48, 4236.[9] S. Cacchi, F. Latorre, G. Palmieri, J. Organomet. Chem. 1984, 268, c48.

[10] A. L. Gottumukkala, J. G. de Vries, A. J. Minnaard, Chem. Eur. J. 2011, 17,3091.

[11] S. Mannathan, S. Raoufmoghaddam, J. N. H. Reek, J. G. de Vries, A. J.Minnaard, ChemCatChem 2015, 7, 3923.

[12] S. Raoufmoghaddam, M. Mannathan, A. J. Minnaard, J. G. de Vries,J. N. H. Reek, Chem. Eur. J. 2015, 21, 18811.

[13] Y. Tsuchiya, Y. Hamashima, M. Sodeoka, Org. Lett. 2006, 8, 4851.[14] B. Q. Ding, Z. F. Zhang, Y. G. Liu, M. Sugiya, T. Imamoto, W. B. Zhang,

Org. Lett. 2013, 15, 3690.[15] W. Q. Kong, Q. Wang, J. P. Zhu, Angew. Chem. Int. Ed. 2017, 56, 3987;

Angew. Chem. 2017, 129, 4045.[16] S. J. Liu, J. R. Zhou, Chem. Commun. 2013, 49, 11758.[17] A. Minatti, X. Zheng, S. L. Buchwald, J. Org. Chem. 2007, 72, 9253.[18] G. Yue, K. Lei, H. Hirao, J. Zhou, Angew. Chem. Int. Ed. 2015, 54, 6531;

Angew. Chem. 2015, 127, 6631.

[19] C. Shen, R.-R. Liu, R.-J. Fan, Y.-L. Li, T.-F. Xu, J.-R. Gao, Y.-X. Jia, J. Am.Chem. Soc. 2015, 137, 4936.

[20] M. Ichikawa, M. Takahashi, S. Aoyagi, C. Kibayashi, J. Am. Chem. Soc.2004, 126, 16553.

[21] S. Cacchi, G. Fabrizi, F. Gasparrini, P. Pace, C. Villani, Synlett 2000, 650.[22] H. Naito, S. Ohsuki, M. Sugimori, R. Atsumi, M. Minami, Y. Nakamura, M.

Ishii, K. Hirotani, E. Kumazawa, A. Ejima, Chem. Pharm. Bull. 2002, 50,453.

[23] U. S. Sørensen, T. J. Bleisch, A. E. Kingston, R. A. Wright, B. G. Johnson,D. D. Schoepp, P. L. Ornstein, Bioorg. Med. Chem. Lett. 2003, 11, 197.

[24] S. Ohira, T. Ida, M. Moritani, T. Hasegawa, J. Chem. Soc. Perkin Trans. 11998, 293.

[25] D. G. Blackmond, Angew. Chem. Int. Ed. 2005, 44, 4302; Angew. Chem.2005, 117, 4374.

[26] Turbomole, 5.8, v. Theoretical Chemistry Group, University of Karlsruhe,Germany, 2002.

[27] A. D. Becke, Phys. Rev. A 1988, 38, 3098.[28] J. P. Perdew, Phys. Rev. B 1986, 33, 8822.[29] A. Sch-fer, H. Horn, R. Ahlrichs, J. Chem. Phys. 1992, 97, 2571.[30] C. Amatore, A. Fuxa, A. Jutand, Chem. Eur. J. 2000, 6, 1474.[31] C. Amatore, A. Jutand, Acc. Chem. Res. 2000, 33, 314.[32] C. Amatore, E. Carre, A. Jutand, Y. Medjour, Organometallics 2002, 21,

4540.[33] I. J. S. Fairlamb, A. R. Kapdi, A. F. Lee, Org. Lett. 2004, 6, 4435.[34] I. J. S. Fairlamb, A. R. Kapdi, A. F. Lee, G. P. McGlacken, F. Weissburger,

A. H. M. de Vries, L. Schmieder-van de Vondervoort, Chem. Eur. J. 2006,12, 8750.

[35] A. Sundermann, O. Uzan, J. M. L. Martin, Chem. Eur. J. 2001, 7, 1703.[36] a) U. Christmann, R. Vilar, Angew. Chem. Int. Ed. 2005, 44, 366; Angew.

Chem. 2005, 117, 370; b) D. A. Culkin, J. F. Hartwig, J. Am. Chem. Soc.2001, 123, 5816; c) D. A. Culkin, J. F. Hartwig, Acc. Chem. Res. 2003, 36,234.

[37] J. A. Casares, P. Espinet, G. Salas, Chem. Eur. J. 2002, 8, 4843.[38] F. Barrios-Landeros, J. F. Hartwig, J. Am. Chem. Soc. 2005, 127, 6944.[39] D. Guest, V. H. Menezes da Silva, A. P. de Lima Batista, S. M. Roe, A. A. C.

Braga, O. Navarro, Organometallics 2015, 34, 2463.[40] S. Mannathan, S. Raoufmoghaddam, J. N. H. Reek, J. G. de Vries, A. J.

Minnaard, ChemCatChem 2017, 9, 551.[41] J. Baker, J. Comput. Chem. 1986, 7, 385.[42] P. H. M. Budzelaar, J. Comput. Chem. 2007, 28, 2226.

Manuscript received: July 24, 2017

Revised manuscript received: August 24, 2017

Accepted manuscript online: September 4, 2017Version of record online: November 30, 2017

ChemCatChem 2018, 10, 266 – 272 www.chemcatchem.org T 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim272

Full Papers