Discovery and Understanding of Transition-Metal-Catalyzed Aromatic Substitution Reactions Transition-Metal-Catalyzedjohn Aromatic Substitution Reactionsf

ACCOUNT 1283

Discovery and Understanding of Transition-Metal-Catalyzed Aromatic Substitution Reactions Transition-Metal-CatalyzedJohn Aromatic Substitution ReactionsF. Hartwig* Department of Chemistry, Yale University, P.O. Box 208107, New Haven, CT 06520-8107, USA Fax +1(203)4326144; E-mail: [email protected] Received 5 January 2006 John F. Hartwig was the recipient of the 2004 Thieme–IUPAC Prize.

amides derived from secondary amines in the presence of Abstract: This article presents studies on the development of cross- coupling reactions of aryl halides and sulfonates with amines, a palladium catalyst containing a sterically hindered alkoxides, and various enolates. Emphasis is placed on the process aromatic phosphine (Equation 1). Although the scope of of developing new catalysts with mechanistic information gained in this prior work was also narrow, it demonstrated that the author’s laboratory. palladium complexes could catalyze aromatic carbon– 1 Introduction nitrogen bond formation. 2 Early Reaction Development Et 3 New Catalyst Developments R L2PdCl2 R Et 4 Mechanism of the Cross-Coupling Processes Bu3Sn N + Br N + Bu3SnBr only for 5 Mechanism of Oxidative Addition Et Et L = (o-Tol)3P Ref. 7,8 6 Reductive Elimination of Amines 7 Stable Palladium Catalysts for Coupling of Primary Amines Equation 1 8 Guidelines for Catalyst Selection Key words: cross-coupling, palladium, amination, arylation, We were interested in developing methods that would mechanism expand the scope of this process and that would occur without tin reagents. Although we did not anticipate that the process would be as general and widely used as it has 1 Introduction become, we did hope that these advances would create a process that is convenient enough to be useful for organic Nucleophilic aromatic substitution is a fundamental synthesis. We hoped to uncover catalysts that would be organic transformation. However, the scope of the direct more active for the amination process and that could ex- reaction is narrow and the temperature for reaction of pand the scope to encompass primary amines, nitrogen unactivated substrates is high. Mild, direct nucleophilic nucleophiles other than amines, and related nucleophiles, aromatic substitution requires highly electron-deficient such as alkoxides. aryl halides. Thus, alternative uncatalyzed methods have We were also interested in uncovering the elementary re- been studied to develop a net aromatic substitution pro- actions of this catalytic process, including the step of the 1 cess, such as reactions through aryne intermediates or catalytic cycle that forms the carbon–nitrogen bond in the 2,3 reactions occurring by radical nucleophilic substitution. product. If the C–N coupling process followed a mecha- Although these reactions are useful in some contexts, the nism that was analogous to that for conventional C–C scope of these stepwise, uncatalyzed methods is also cross-coupling, then the carbon–nitrogen bond would be narrow and the tolerance for accompanying functional formed by a type of reductive elimination that was un- groups is poor. known when we started our work.9,10 We sought to use My group has sought transition-metal catalysts for nu- this type of mechanistic information on the catalytic pro- cleophilic aromatic substitution reactions that occur with cess to design or select catalysts that would give rise to the both heteroatom nucleophiles and common enolate desired increase in activity and scope of the process. 4–6 nucleophiles. As a result of these efforts, we have Due to the efforts of students and postdocs in my group, developed mild methods to prepare aromatic amines, as well as work by many other groups, including the ex- ethers, sulfides, and a-aryl carbonyl compounds with tensive work of Steve Buchwald’s group at MIT,11,12 the phosphine-ligated palladium catalysts. Our work on scope of this reaction is now extremely broad. Equation 2 metal-catalyzed aromatic substitution, or ‘cross-cou- summarizes the types of catalysts now employed in the pling’, stemmed from work by Kameyama, Kosugi and amination process and an overall view of the scope of the 7,8 Migita from the early 1980’s. These authors reported C–N coupling. With these catalysts, the amination reac- the coupling of electron-neutral aryl bromides with tin tions can be conducted with low loadings of catalyst, and can now be conducted with aromatic bromides, chlorides, SYNLETT 2006, No. 9, pp 1283–1294 01.06.2006 triflates, in some cases with aryl iodides, and most recent- Advanced online publication: 22.05.2006 ly with aryl tosylates. Both primary and secondary amines DOI: 10.1055/s-2006-939728; Art ID: A40606ST © Georg Thieme Verlag Stuttgart · New York undergo this reaction, although the optimal catalyst for 1284 J. F. Hartwig ACCOUNT these two classes of amines is often different. The reaction 2 Early Reaction Development has been adopted for many types of applications, such as the synthesis of aromatic amines and nitrogen hetero- The first synthetic advance was published concurrently by cycles for biological and medicinal applications, synthesis Steve Buchwald’s group and by us13,14 and showed that of sophisticated materials for electronic applications, and the reaction could be conducted with amine nucleophiles synthesis of anionic nitrogen ligands for applications in in the presence of an alkoxide or silylamide base with the both asymmetric catalysis and the synthesis of polyole- catalyst originally used by Kosugi and Migita for the fins. coupling of tin amides. A wide variety of catalysts have now been evaluated for this reaction, and the current Pd/2 L R scope of this reaction is summarized in a broad sense in X X = Cl, Br, I, OTf, OTs N base R' Equation 2. In addition to amines, some other nitrogen + HNRR' 25–80 °C nucleophiles that are useful in synthetic applications Y Y L = hindered monodentate ligands undergo the coupling process with broad scope. For

P(o-Tol)3, P(t-Bu)3, Ph5FcP(t-Bu)2 (Q-phos), heterocyclic carbenes, example, the reaction of ammonia equivalents, such as – 15,16 (Biaryl)PR2, OP(t-Bu)2, Solvias metallacycle, Verkade's phosphatranes benzophenone imine and lithium hexamethyldisilyl- or chelating bidentate ligands amide (LiHMDS),17,18 occur in good yield with many DPPF, BINAP, Josiphos ligands, Xantphos aromatic halides. Reactions of benzophenone imines and Broad scope: cyclic secondary, acyclic secondary, primary aliphatic, aromatic amines an alkoxide base are milder than those with the more NH strongly basic LiHMDS and can be conducted with ortho- LiN(SiMe ) H N-N=CPh Ph Ph 3 2 2 2 substituted aryl halides, but the hydrolysis of the aromatic Narrower scope: silylamide products is more facile. Benzophenone hydra- H O O O O Boc N HN O HN zone also reacts with aryl halides with broad scope, and S S H N R R R H N O-t-Bu H N p-Tol HN 2 2 2 Boc the products from these reactions are precursors to a variety of nitrogen heterocycles.19–22 Equation 2

Biographical Sketch

John F. Hartwig was born transition-metal complexes. In addition to the Thieme– in 1964 outside of Chicago, He has developed a selec- IUPAC Prize in Synthetic and was raised in upstate tive catalytic functionaliza- Organic Chemistry, Pro- New York. He received a tion of alkanes, a method for fessor Hartwig received the B.A. in 1986 from Princeton formation of arylamines and Leo Hendrik Baekeland University, and a Ph.D. in aryl ethers from aryl halides Award in 2003, the A.C. 1990 from the University of or sulfonates, a method for Cope Scholar Award in California, Berkeley under the direct conversion of car- 1998, the Camille Dreyfus the collaborative direction bonyl compounds to a-aryl Teacher-Scholar Award in of Robert Bergman and carbonyl derivatives, a 1997, a Union Carbide Richard Andersen. After an system for the catalytic ad- Innovative Recognition American Cancer Society dition of amines to vinylare- Award in 1995 and 1996, postdoctoral fellowship nes and dienes, and highly the National Science Foun- with Stephen Lippard, he selective catalysts for the dation Young Investigator began an appointment at regio- and enantioselective Award in 1994, and both Yale University in 1992, amination of allylic carbon- Dupont and Dreyfus Foun- where he is now the Irénée ates. With each system, his dation New Faculty Awards P. duPont Professor of group has conducted exten- in 1992. He was the 1998 Chemistry. He and his wife sive mechanistic investiga- R.C. Fusan lecturer at the Anne Baranger have recent- tions. He has revealed University of Illinois, the ly accepted positions at the several new classes of re- first AstraZeneca lecturer in University of Illinois, where ductive eliminations, has Stockholm, and the 2001 they will move during 2006 isolated discrete compounds Carl Ziegler lecturer at Mül- with their two daughters that functionalize alkanes, heim. Most recently, he was Pauline and Amelia. and has reported unusual named the 2006 recipient of Professor Hartwig’s re- three-coordinate arylpalla- the ACS Award in Organo- search focuses on the dis- dium complexes that are metallic Chemistry. covery and understanding of intermediates in cross-cou- new reactions catalyzed by pling.

Synlett 2006, No. 9, 1283–1294 © Thieme Stuttgart · New York ACCOUNT Transition-Metal-Catalyzed Aromatic Substitution Reactions 1285

Several other nitrogen nucleophiles that participate in the X O O Pd/L Y Y + R'' coupling process are listed at the bottom of Equation 2. R base R The scope of the coupling of these nucleophiles tends to X = Br, Cl R' R' R'' be narrower, the amount of catalyst required for good L = bidentate or hindered monodentate ligands yields tends to be higher, and the temperatures required BINAP, DTPF, Xantphos i-Pr i-Pr P(t-Bu)3, PAr(t-Bu)2, NN for full conversion tend to be higher than those of reac- 1-Ad2P(t-Bu)2 tions of amines. Although these substrates have diverse Ph5FcP(t-Bu)2 i-Pri-Pr structures, their common feature is a nitrogen that is less suitable enolate include those from: basic and less nucleophilic. Thus, the transition-metal- O O O Br R'' R'' R'' O catalyzed coupling process occurs with much broader R RO R2N R'' R' R' R' N scope than the uncatalyzed aromatic substitution but the R R' R' O O O O NC reaction does remain somewhat sensitive to the electronic CN N=CPh2 properties of the nitrogen nucleophile. R RO RO OR RO Consistent with this trend, the couplings to form carbon– Equation 3 oxygen bonds from aryl halides and alkoxides is challenging. Initially my group and Steve Buchwald’s group pub- 3 New Catalyst Developments lished the coupling of alkoxides with electron-poor aromatic halides with catalysts containing aromatic bis- Substantial effort in my group has focused on using mech- phosphines such as BINAP [2,2'-bis(diphenylphosphino)- anistic principles to uncover improved catalysts for these 1,1'-binaphthyl] and DPPF [1,1-bis(diphenylphosphi- coupling reactions. This approach led us to begin using na)ferrocene].23–27 More recently, this process has been bisphosphines, such as DPPF, for the catalytic process.42 reported with aryl halides that are electron-neutral and In 1998, researchers at Tosoh Company in Japan reported lack activating groups.28–35 These etherifications have that P(t-Bu) catalyzed the amination of aryl halides with been accomplished with phosphine ligands that are ex- 3 relatively high turnover numbers in refluxing xylenes.43 tremely hindered. For example, we have reported the con- As we began to explore the mechanism of this reaction, version of aryl halides to diaryl ethers and tert-butyl we quickly found that reactions conducted with a 1:1 ratio ethers with a ligand called Q-phos that contains five phen- of P(t-Bu) to Pd(dba) occurred rapidly at room tempera- yl groups on one cyclopentadienyl ring of a di-tert- 3 2 ture.44 butylphosphino ferrocene unit.29 Buchwald has reported the coupling of aryl halides with aryloxides and alkox- Scheme 1 summarizes our contribution to the amination ides,31,35 including those with hydrogens a to oxygen, with of aryl halides with this ligand system. Reactions of catalysts generated from biaryldialkyl phosphine ligands. secondary alkylamines, primary aromatic amines, or secondary aromatic amines occurred in good yields at We have also developed the coupling of carbon nucleo- room temperature with a variety of bromoarenes. Like- philes that can be generated readily by deprotonation of a wise, the reaction of aryl chlorides occurred under condi- carbonyl compound or a nitrile. After obtaining a seren- tions that were milder than those with most catalysts dipitous preliminary result, we appreciated that the similar reported previously, and the reaction with aniline was pKa values of carbonyl compounds and amines could al- even accomplished at room temperature. However, this low ketones, esters, amides and nitriles to undergo cou- catalyst system is not particularly effective for the reac- pling with aryl halides under conditions that are analogous tions of primary aliphatic amines. to those of the coupling of amines, and we began to develop the a-arylation of carbonyl compounds.36,37 Now, ke- 1% Pd(dba) / Br 2 NR2 tones, esters, amides, nitriles, 1,3-dicarbonyl compounds 1% P(t-Bu) R + 3 R and cyano esters react with aryl halides and triflates in the HNR'2 + NaO-t-Bu r.t., 15 min–6 h 6 83–97% presence of an appropriate base and palladium catalyst. R = donating or Buchwald reported early examples of this reaction with withdrawing; o, m or p 37 1% Pd(dba) / BINAP as ligand and we reported examples with a Br 2 NHPh 1% P(t-Bu)3 sterically hindered version of 1,1¢-bis(di-o-tolylphos- + H2NPh 36 r.t., 1 h phino)ferrocene (DTPF). Improved catalysts have + NaO-t-Bu 97% been reported by various groups,38–40 including ours,41 with sterically hindered alkyl phosphines. Complexes Cl 1–2% Pd(dba)2/ NR'2 R + 1–2% P(t-Bu) generated from this class of ligands catalyze the reactions HNR'2 3 R + NaO-t-Bu 25–80 °C, 4–24 h of the carbonyl compounds and nitriles shown at the R' = alkyl, aryl, H 76–97% bottom of Equation 3. Scheme 1

Considering that this catalyst contained a 1:1 ratio of P(t-Bu)3 to palladium, a complex that already contains a 1:1 ratio of palladium to P(t-Bu)3 that can rapidly undergo

Synlett 2006, No. 9, 1283–1294 © Thieme Stuttgart · New York 1286 J. F. Hartwig ACCOUNT

Ph FcP(t-Bu) = reduction to palladium(0) could lead to even faster 1 mol% Pd(dba)2 5 2 Cl R reactions. Indeed, the dimeric palladium(I) complex in 2 mol% Ph5Fc(t-Bu)2 HNRR' Y + H N Y P(t-Bu)2 Ph Fe Equation 4 is spectacularly reactive for the amination of R' NaO-t-Bu, toluene Ph 45 70–100 °C Ph Ph aryl chlorides and bromides. The adamantyl di-tert- Ph butylphosphine complex is more stable toward air and cat- O Bu Me "Q-phos" H alyzes these processes with slightly higher yields than the N N N N Bu Ph Ph P(t-Bu)3 analogue. Yet, both of these complexes lead to the coupling of secondary aliphatic amines and aryl- MeO 95% MeO 93% 95% MeO 95% amines with aryl chlorides within 15 min at room temper- hexyl hexyl hexyl CH2Ph ature with 0.5 mol% of the dimer. Of course, aryl NH NH NH NH bromides react as fast as aryl chlorides. Like reactions of 86% MeO primary amines catalyzed by the 1:1 ratio of P(t-Bu)3 and Me 94% Me MeO2C 91% 92% Pd(dba) , reactions of primary amines catalyzed by this mono:diarylation (K3PO4) 2 = 45:1 dimer occurred in low yields. Equation 5 0.5 mol% Br Cl R(t-Bu)2P Pd Pd P(t-Bu)2R NR (R = t-Bu, 1–Ad) 2 R Br R + HNR2 duces the coupling of primary amines with aryl and (R = alkyl, aryl) r.t., 15 min heteroaryl chlorides with remarkably high turnover + NaO-t-Bu numbers46 and with aryl tosylates under unusually mild R = donating or 47 withdrawing; o, m or p conditions. A short summary of the reactions of aryl tosylates are shown in Scheme 2. Reactions of aryl and Equation 4 heteroaryl chlorides will be presented in detail at the close of this Account. My group has also developed an extremely hindered fer- rocenyl ligand that generates catalysts that react with broad scope for a variety of cross-coupling reactions.30 1% PdCl2(PhCN)2 + P 2 Equation 5 presents the coupling of amines with aryl Fe PCy2 OTs NHR' chlorides in the presence of catalysts containing this (CyPF-t-Bu) pentaphenylferrocenyl di-tert-butyl phosphine, which we Y Y r.t. or 80 °C call Q-phos after Quinetta Shelby who first prepared it. Complexes of this ligand react slightly slower than those + H2NR' + NaOt-Bu 50 ppm 1% NHR' of the tri-tert-alkylphosphines for the amination of aryl Cl Pd(OAc) + CyPF-t-Bu chlorides. However, complexes of this ligand catalyze the Y Y r.t. to 100 °C coupling of a broad scope of amines and aryl chlorides in high yields. As shown in Equation 5, complexes of this + H2NR' + NaOt-Bu 10 ppm 1% ligand catalyze the coupling of cyclic or acyclic secondary Pd(OAc) + CyPF-t-Bu amines, including the difficult coupling of an acyclic Y Cl Y NHR' N N secondary amine with an electron-rich aryl chloride. In r.t. to 100 °C addition, this complex catalyzes the coupling of both aro- + H2NR' + NaOt-Bu matic and aliphatic primary amines. The coupling of an Scheme 2 unhindered aryl chloride with an unhindered amine without an excess of the primary amine is a difficult reaction to conduct with high selectivity because the product Many of the factors that create an active catalyst for the secondary amine can undergo arylation to generate a di- amination of aryl halides also create an active catalyst for aryl alkylamine. However, complexes ligated by Q-phos the a-arylation of carbonyl compounds. For example, the catalyze the selective monoarylation of a primary amine coupling of ketone enolates with aryl chlorides and in high yield. bromides occurs under mild conditions in the presence of catalysts generated from a 1:1 ratio of P(t-Bu)3 and Finally, my group has uncovered palladium catalysts with 41 Pd(OAc)2 (Equation 6). Aryl bromides of varying elec- a hindered bisphosphine that addresses many of the limi- tronic and steric properties couple with acetophenone de- tations on the reactions of aryl halides and sulfonates with rivatives and two equivalents of base to generate at room primary amines. For reasons that will be discussed later in temperature the monoarylation product. Reactions with this manuscript, complexes with sterically hindered one equivalents of base occur in high yield at the methyl- bidentate ligands are highly effective for the coupling of ene positions of cyclic ketones, such as cyclohexanone, at primary amines with aryl chlorides, heteroaryl chlorides, the methylene positions of acyclic ketones, such as pro- and aryl tosylates. With a large bite angle, rigid backbone piophenone, and at the methine positions of ketones, such and sterically demanding alkyl substituents, the Josiphos as isobutyrophenone. When two types of enolizable ligand containing one dicyclohexylphosphino and one di- hydrogens are present in the ketone, the reaction occurs tert-butylphosphino group generates a catalyst that in-

Synlett 2006, No. 9, 1283–1294 © Thieme Stuttgart · New York ACCOUNT Transition-Metal-Catalyzed Aromatic Substitution Reactions 1287 selectively at the less hindered of the two enolizable posi- of these reactions encompass aryl halides with a range of tions. Although our initially published work reported only electronic properties, but the reactions of pyridyl halides a 3:1 ratio for arylation at the methylene position of ethyl with a,a-disubstituted esters occurred in high yield. For isopropyl ketone, a change in solvent from THF to toluene reasons we do not yet understand, the reactions of pyridyl led to a 12:1 ratio of these products. Further, changing the halides with alkali metal enolates of acetates and propi- ligand from tri-tert-butyl phosphine to the sterically hin- onates did not occur. dered, saturated N-heterocyclic carbene led to coupling of phenyl chloride with the same ketone to generate a single 0.2–2% Pd(dba)2/ O carbene or P(t-Bu)3 O product at room temperature in 91% isolated yield. Reac- R' Y R' X + R'' tions of aryl chlorides with the same types of ketones OR NaN(SiMe3)2, OR R'' or LiNCy2 Ar occurred with a 1:1 ratio of Pd(OAc)2 and P(t-Bu)3 in high r.t., 12 h, toluene R' = H, Me; R = t-Bu yield after roughly 12 hours of reaction time at 70 °C. R' = Me, Et, Ph; R = Me, Et, Bn O O Y P(t-Bu)3/Pd(OAc)2 R R Equation 7 X + R'' R'' NaO-t-Bu Ar R' 25–70 °C, THF R' X = Br, Cl Although the scope of these reactions encompasses aryl Equation 6 halides with a range of electronic properties, the basicity of the enolates limits the type of functional groups that are a The coupling of aryl halides with the enolates of esters is tolerated by this -arylation chemistry. For example, reac- challenging because these enolates are less stable than tions of substrates with electrophilic functionality, such as those of ketones.48,49 Even reactions of aryl halides with carbonyl, cyano and nitro groups, as well as substrates with enolizable hydrogens, did not occur in synthetically the lithium enolates of tert-butyl acetate and tert-butyl useful yields. Thus, we sought reactions of zinc and sili- propionate, which Rathke showed to be more stable than con enolates to improve functional group tolerance. the enolates of less hindered esters,49 occurred in modest yields with catalysts that required elevated temperatures. The use of zinc enolates has provided one solution to the However, the high reactivity of catalysts generated from problem of functional group tolerance in the reactions of 51 P(t-Bu)3 and N-heterocyclic carbenes allows the catalytic enolates with aryl halides, as summarized in Equation 8. coupling to occur much faster than the decomposition of The reaction of a Reformatsky reagent, either isolated or the ester enolates.17,50 As summarized in Equation 7, The generated in situ from the a-bromo ester and activated reaction of tert-butyl acetate and tert-butyl propionate zinc metal, occurred with aryl halides possessing an array with aryl bromides possessing a variety of electronic of functionality that was not tolerated by the reactions of properties occurred in high yields. Reactions of tert-butyl alkali metal enolates. For example, the reactions of bro- acetate occurred fastest and in the highest yields with lith- mobenzophenone, bromo-4-methylbenzoate, 4-bromoni- ium dicyclohexylamide as base and a 1:1 combination of trobenzene and bromobenzonitriles all occurred in good 50 P(t-Bu)3 and Pd(dba)2 as catalyst. Reactions of tert-butyl yield with the zinc enolate generated from a-bromo tert- propionate occurred in the highest yields with the N-het- butyl acetate and tert-butyl propionate. These reactions erocyclic carbene as ligand.17,50 Reactions of methyl, occurred in the highest yields with the lowest catalyst ethyl, and benzyl esters that are disubstituted in the a-po- loadings when the palladium catalyst was generated from sition occurred in high yields with LiNCy2 as base and Pd(dba)2 and Q-phos. 50 P(t-Bu)3 and Pd(dba)2 as catalyst. Not only did the scope

Y O O Br + BrZn Ar P(t-Bu)2 Ot-Bu 1–2% Pd(dba)2/Q-phos OR Ph Fe Ph R R Ph Ph r.t., THF Ph R = H, Me R = H, Me "Q-phos" O O O N O MeO O 2 O O O O-t-Bu O-t-Bu O-t-Bu O-t-Bu O-t-Bu R R R CN R NO2 R R = H 75% R = Me 87% R = H 96% R = H 91% R = H 87% R = Me 89%, 70 °C R = Me 85%

N HO H N O N O O O 2 O O-t-Bu O-t-Bu O-t-Bu O-t-Bu O-t-Bu R OH R R R = H 90% R = H 79% R = H 91% R = H 83% R = H 80% R = Me 91% R = Me 95% R = Me 67%

Equation 8

Synlett 2006, No. 9, 1283–1294 © Thieme Stuttgart · New York 1288 J. F. Hartwig ACCOUNT

As shown in Equation 8, the scope of the a-arylation of O R1 OTMS 1% Pd(dba) , 2% P(t-Bu) 2 3 Ar ester enolates also included aryl halides with protic func- OR2 tionality and bromopyridines. The enolate of tert-butyl Me OR2 0.5 equiv ZnF2, ArBr, DMF, 80 °C, 12 h Me R1 1 2 acetate reacted with 3- and 4-bromopyridines in good R = H, R = t-Bu including: O R = OMe H 80% yields, and the enolate of tert-butylpropionate reacted R1 = Me, R2 = Me R = OMe Me 94% 1 2 R R = Me H 68% with 3-bromopyridine in good yield. The zinc enolate of R = OBn, R = Me, Et, t-Bu R = Me Me 78% Br R = Ph Me 99% tert-butyl acetate and tert-butylpropionate also reacted with 2- and 4-bromophenol and 4-bromoaniline. These O OSiMe3 O O O O Pd(dba)2, Pt-Bu3 reactions were accomplished with either two equivalents O N O N O N PhBr, additive, DMF + of the enolates or one equivalent of KH, followed by one Ar Ar equivalent of the enolate. R major R minor R

The coupling of aryl halides with the alkali metal enolates R additive temp (°C) dr (yield of major) of amides occurred in synthetically useful yields in some ZnF2 80 88:12 (67%) cases, but the scope and typical reaction yields were much Zn(Ot-Bu)2 r.t. 91:9 (70%) narrower than those of the reactions of alkali metal eno- 95:5 (61%) lates of ketones and esters.52 Thus, we have recently rein- Zn(Ot-Bu)2 r.t. vestigated the a-arylation of amides, and have found that Scheme 3 the reactions of zinc enolates of amides occur in higher yields (Equation 9).51 Although we do not have a precise explanation for the lower yields and narrow scope for the The stereochemistry of the carbon a to the carbonyl group reactivity of the alkali metal enolates of amides, the high in products generated from the silicon enolates are stable basicity of the reaction medium and the relative pKa to the reaction conditions because of the low basicity of values of acetamides and a-aryl acetamides lead to de- these enolates.53 As shown at the bottom of Scheme 3, the composition of the catalyst and to competing formation of silyl ketimine that bears an Evans auxiliary reacted with products from diarylation. In contrast, the reactions of the aryl halides diastereoselectively in the presence of the pal- zinc enolates of acetamides occur in high yields with ladium catalyst. Further, the ratio of diastereomers results broad functional-group tolerance and without formation from a kinetic selectivity. The ratio of diastereomers was of products from diarylation. Thus, the zinc enolate of di- constant throughout the reaction, and no epimerization of ethyl acetamide and diethyl propionamide (Equation 9) the pure diastereomer of a related compound, which was occurred in good yields, not only with an electron-neutral added to the reaction as a probe of the basicity of the aryl bromide such as 4-bromo-tert-butyl benzene, but medium, occurred. Further, trimethylsilyl enolates of the with aryl bromides that failed to react with the lithium reagents developed by Ley react with aryl halides to gen- enolates of amides, such as bromobenzoates, bromo- erate products of very high diastereomeric purity. nitroarenes, and bromobenzonitriles.

O O 4 Mechanism of the Cross-Coupling Processes Y BrZn 1–2% Pd(dba)2/Q-phos Ar Br + NEt2 NEt2 Catalytic cross-coupling processes typically occur in R r.t., dioxane R R = H, Me R = H, Me three stages: oxidative addition of an aryl halide to a palladium(0) complex, reaction of the aryl halide complex Equation 9 with a nucleophile to generate an intermediate with the nucleophile bound to palladium, and reductive elimina- Clearly, silicon enolates could react with even higher tion to couple the aryl group and the nucleophile and to functional group tolerance than zinc enolates, and the eno- regenerate the palladium(0) complex. Each of these three lates could be formed directly from the carbonyl com- stages is a multi-step process often involving ligand dis- pound. Xiaoxiang Liu surmised that addition of zinc sociation prior to the elementary oxidative addition, trans- fluoride would transform the silicon enolates to zinc eno- metalation and reductive elimination steps. lates, which might react as just described. The origin of Scheme 4 highlights several features of this general mech- the effect of the zinc fluoride or zinc tert-butoxide addi- anism that should be considered when rationalizing reactive is likely to be different than that originally proposed, tivity or when designing or selecting ligands for cross- but this combination of silicon enolate and zinc additive coupling. First, the rates of reactions of aryl chlorides and did react with aryl halides in the presence of palladium of most reactions of aryl bromides are controlled by the 51,53 and P(t-Bu)3 as catalyst, as summarized in Scheme 3. rate of oxidative addition. Thus, the phosphine-ligated Unlike the reactions of alkali metal enolates, these reac- palladium(0) complex is the catalyst resting state,54 and tions occurred in highest yields in a polar solvent such as efforts to accelerate the rates of the cross-coupling chem- DMF at 80 °C over 12 h. In addition to improving the tol- istry must focus on increasing the rate of oxidative addi- erance of functional groups on the aryl halide, this proce- tion. However, the yields and scope of the reaction are dure allowed the a-arylation of a-alkoxy esters that failed generally controlled by the reductive elimination. The to couple, as their alkali metal enolates, with aryl halides. reductive elimination step must be faster than side reac-

Synlett 2006, No. 9, 1283–1294 © Thieme Stuttgart · New York ACCOUNT Transition-Metal-Catalyzed Aromatic Substitution Reactions 1289 tions such as b-hydrogen elimination, protonolysis of the the chelating ligand to generate a bent two-coordinate metal amide, alkoxide or enolate intermediate or dispro- complex prior to oxidative addition.59 A minor pathway portionation of the aryl palladium complexes to form bi- for addition to [Pd(BINAP)2] and presumably [Pd(DP- 2 aryl complexes that undergo reductively elimination of PF)2] occurs by reaction of the aryl halide with [Pd(k -bis- biaryl. We have studied the mechanism of each of the phosphine)(k1-bisphosphine)], most likely by dis- three stages of the catalytic amination of aryl halides and placement of the k1-ligand by the aryl halide to generate a-arylation of carbonyl compounds. [Pd(BINAP)(ArX)] or [Pd(DPPF)(ArX)] that precedes oxidative addition. The bent two-coordinate palladium(0) Catalyst complexes are less stable than the linear species and, in Resting State the bent conformation, project orbitals that are appropriate Turnover- R L Pd L to interact with the incoming aryl halides. Limiting Step N Br R' + L – L Others have suggested that the amine coordinates to the Key for Selectivity, L Pd palladium(0) species and that the amine complex under- Yield, and Scope goes oxidative addition faster than the two-coordinate 60 L L palladium(0) species. If so, then the order in which the nucleophile and aryl electrophile react in the amination of

Ph Ph aryl halides would be different from the order in which LPd LPd NRR' Br they react in other types of cross-coupling, and previous conclusions about the species that add haloarenes would require reconsideration. This assertion was based on base·HBr HNRR' kinetic modeling of rate data obtained by calorimetry on + base reactions conducted with a catalyst generated in situ. Scheme 4 During the past few years, we have shown that the rate of oxidative addition to [Pd(BINAP)2] is much less than first order in amine,61 and that the catalytic reaction shows no 5 Mechanism of Oxidative Addition positive order in amine.62 More than two years after the original submission of these data, the results of this work Scheme 5 summarizes the mechanism of oxidative addi- have been published.61,62 The apparent order in amine tion of aryl halides to several phosphine-ligated Pd(0) measured under the conditions of the catalytic system re- complexes. Fauvarque and Amatore55,56 independently sulted from catalyst decomposition. showed that oxidative addition of aryl halides to tri- The effect of steric hindrance on the oxidative addition of phenylphosphine-ligated palladium(0) occurs after dis- aryl halides is summarized in the energy diagram in sociation of ligand from (PPh ) Pd to generate linear 3 3 Figure 1. The unsaturated reactive intermediate lies uphill (PPh ) Pd. This intermediate then undergoes the cleav- 3 2 from the starting palladium(0) complex, but reaction of age of the carbon–halogen bond to generate this intermediate with the aryl halide generates a stable (PPh ) Pd(Ar)(X). We showed that palladium(0) com- 3 2 arylpalladium halide complex. The largest difference be- plexes with extremely hindered phosphines, such as P(o- tween the starting palladium(0) species and the reactive Tol) or Q-phos, which are stable in their two-coordinate 3 intermediate is the coordination number. Thus the pres- geometry, undergo oxidative addition of aryl halides after ence of more hindered ligands on the palladium(0) com- dissociation of phosphine.57,58 The three-coordinate plex will increase the energy of the ground state more than product from this oxidative addition then dimerizes to they will increase the energy of the low-coordinate palla- form the stable product when the ligand is P(o-Tol) . 3 dium(0) intermediate. As a result, the difference in energy between the ground state and the reactive intermediate L L [ArBr] Ar will be smaller when more hindered ligands are bound to Pd L LPdL Pd L L Br the palladium(0) than when less hindered ligands are

L = PPh3 bound to the palladium(0). L [ArBr] Ar L Br L Pd L L Pd Pd Pd Two examples presented in Scheme 6 demonstrate the Br Ar dramatic acceleration of the rate of oxidative addition by L = P(o-Tol)3 2 complexes of hindered monodentate ligands. First, the L L L [ArBr] L Ar oxidative additions of aryl bromides to palladium(0) Pd Pd Pd L L L L Br ligated by tri-tert-butylphosphine are much faster than 63,64 L = BINAP, DPPF additions to those with less hindered ligands. For example the oxidative addition of phenyl bromide to a Scheme 5 combination of Pd(dba)2 and P(t-Bu)3 occurs at room temperature within minutes, whereas the oxidative We have also shown that the major pathway for oxidative addition of phenyl bromide to the bisphosphine-ligated addition to palladium(0) complexes of bis-phosphines, palladium(0) requires 70 °C and several hours.65 The such as BINAP and DPPF, occurs by full dissociation of products from oxidative addition of phenyl bromide to

Synlett 2006, No. 9, 1283–1294 © Thieme Stuttgart · New York 1290 J. F. Hartwig ACCOUNT

Ph Ph Tol KN(tolyl)2 80 °C + (PPh ) Pd Ph3P Pd PPh3 Ph3P Pd PPh3 N 3 4 PPh3 Tol I N Tol Tol 90% Increased size of Kinetic Studies Show Two Concurrent Mechanisms L L Pd L Ph K1 Ph k 3-coord: Tol2N 1 L Pd Pd NPh(Tol)2 + LPd or + L Tol2N L L L Pd L Ph K2 L NTol2 k 4-coord: 2 Oxidative addition Pd Pd L2Pd + NPh(Tol)2 L or Tol2N L L Ph L or LAr Pd L L Pd L Pd Pd L Ar X L X Scheme 7 L Figure 1 To confirm the kinetic data that implied that reductive elimination occurred from four-coordinate complexes, we the combination of Pd(dba)2 and P(t-Bu)3 have unusual prepared a series of arylpalladium complexes ligated by coordination geometries with only three heavy atoms bisphosphines. The DPPF complexes underwent reduc- 63,64 bound to the metal. tive elimination of aryl amines in good yields42 and, for Second, reactions of phenyl tosylate with the palladium(0) this reason, we began to investigate coupling of amines complexes of a very hindered Josiphos ligand with one di- with aryl halides catalyzed by palladium complexes bear- tert-butylphosphino and one dicyclohexylphosphino ing chelating phosphines. The studies on the reductive group (CyPF-t-Bu) occur within minutes at room temper- eliminations to form C–N bonds summarized in Scheme 8 ature.47,66 Considering that no oxidative additions of aryl revealed several principles. tosylates have been published previously and previous First, the clean formation of amine confirmed that C–N palladium-catalyzed couplings of aryl tosylates have re- bond-forming reductive elimination can occur from a quired activated substrates or high temperatures, the four-coordinate palladium complex. Second, the relative room-temperature oxidative addition of phenyl tosylate rates of reactions of complexes with different amido by the palladium(0) complex is remarkable. groups showed that reductive elimination was faster from complexes with more electron-donating groups on nitrogen than from complexes with more electron-withdraw- 6 Reductive Elimination of Amines ing groups on nitrogen.42,68 For example, the reductive elimination of N-alkylarylamines from an alkyl amido As noted in Scheme 4, reductive elimination controls the complex occurred over several hours at 0 °C, whereas the scope and yields of the catalytic amination of aryl halides. reductive elimination of triaryl amine from a ditolylamido When my group began studying the coupling of amines complex required several hours at 65 °C and the reductive with aryl halides, no examples of reductive elimination of elimination of the benzamidate complex did not occur amines from a transition metal complex were known. Yet, under any conditions. Third, reductive elimination was today, this reaction occurs with many types of anionic faster from complexes with more electron-poor aryl nitrogen ligands and with a broad scope of aryl and vinyl groups bound to palladium.68 Taken together with the groups. studies of the electronic effects of substituents at nitrogen, Initially, we studied the reductive elimination of aryl- these studies showed that the pairing of an electron-rich amine from arylpalladium amido complexes ligated by nitrogen ligand and an electron-poor aryl group leads to triphenylphosphine (Scheme 7). As one representative the fastest reductive elimination. Although the reason for example, arylpalladium diarylamido compounds under- this effect is likely to be complex, one can consider that went reductive elimination of triarylamines in good yields this trend could result from a favorable pairing of a nu- at 80 °C.67,68 Kinetic studies of the reaction of this com- cleophilic nitrogen ligand and an electrophilic aryl group. plex and related anilido and alkylamido complexes Finally, arylpalladium amido complexes with chelating showed that reductive elimination occurred by two con- ligands undergo reductive elimination faster than they un- current pathways: fast reductive elimination from a three- dergo b-hydrogen elimination.42,68 We showed that b-hy- coordinate arylpalladium amido complex and slower re- drogen elimination from square-planar amido complexes ductive elimination from a cis four-coordinate complex. occurs after the generation of a three-coordinate amido

(t-Bu) neat ArBr (t-Bu)2 2 Br P L P PhOTs OTs Pd(dba)2 Ph Pd Pd L Pd r.t., 5 min benzene Ph L P P Cy Cy2 r.t., 5 min 2 CpFe CpFe L = P(t-Bu)3, P(1-Ad)(t-Bu)2, or Q-phos L = P(o-Tol)3 92% Scheme 6

Synlett 2006, No. 9, 1283–1294 © Thieme Stuttgart · New York ACCOUNT Transition-Metal-Catalyzed Aromatic Substitution Reactions 1291

69,70 complexes. Thus, complexes with chelating bidentate Ar Bu3SnNEt2 Ar Ar LnPd LnPd LnPd Ar–H ligands undergo slow b-hydrogen elimination, but under- Br NEt2 H go reasonably facile reductive elimination. This retarda- Ar–NEt2:Ar–H Ar–NEt tion of the rate of b-hydrogen elimination leads to better L = P(o-Tol)3 9:1 2 P(o-Tol)2Ph 8:1 selectivity in the catalytic process for formation of aryl P(o-Tol)Ph2 3:1 amines over the formation of arenes by the combination of PPh3 1:1.2 b-hydrogen elimination of the amide to generate a palladi- Scheme 9 um hydride and reductive elimination of arene. reductive elimination of triarylamine over about two 72 Ph 0 °C hours at –10 °C. Analogous complexes ligated by DPPF (DPPF)Pd PhNH(i-Bu) or by triphenylphosphine required temperatures in the N 64% H range of 80 °C to undergo reductive elimination on the Ph 25 °C same time scale. HNPh2 (DPPF)Pd 1.5 h NHPh 80% Ar Tol THF Tol Ph t-Bu3P Pd N MeO N + Pd[P(t-Bu)3]2 85 °C –10 °C, 2 h (DPPF)Pd PhN(tolyl)2 Tol Tol P(t-Bu)3 N(tolyl)2 1.5 h 90% 95% compare to 80 °C, 2 h for DPPF complexes Ph 110 °C O O (DPPF)Pd X Ph Equation 10 N Ph N H H Ph Scheme 8 We have conducted related studies on the reductive elimination from a series of arylpalladium alkyl complexes containing alkyl groups with varied electronic proper- Results presented in the early sections of this paper show ties (Equation 11).73,74 These reactions are much less that the scope of the coupling to form C–N and C–O bonds sensitive to electronic effects than the reductive elimina- is broader with catalysts containing sterically hindered tion to form carbon–oxygen or carbon–nitrogen bonds. phosphines than with complexes containing less hindered For example, the aryl palladium complexes of enolates ligands. The origin of this effect can be interpreted in derived from amides, esters and ketones underwent reduc- terms of the energy diagram in Figure 1. The reductive tive elimination with rates that were indistinguishable. elimination reaction begins with a four- or three-coordi- These studies show that the reductive elimination to form nate palladium complex and generates the combination of a carbon–carbon bond involves a transition state that is amine and a two- or one-coordinate intermediate that is less polar than the reductive elimination to form a carbon– likely to be higher in energy than the starting arylpalladi- nitrogen or carbon–oxygen bond. We also showed that the um amide complex. Trapping of the unsaturated palladi- stability of the main group metal enolates more likely con- um(0) intermediate with phosphine would then generate a trols the yields of these reactions than the chemistry of the stable palladium(0) species. Sterically hindered ligands palladium enolates. Moreover, we showed that the selec- on the metal center will, again, increase the energy of the tivity for reaction at the less hindered position of a ketone reactant more than they increase the energy of the low-co- with two different types of enolizable hydrogens is de- ordinate reactive intermediate. Therefore, reductive elim- rived from the greater thermodynamic stability of the ination occurs faster from complexes with more hindered complex of the less hindered enolate. Because the rates of ligands. reductive elimination from the two enolates are similar, Sterically hindered ligands also enhance the relative rates the ratio of products is controlled predominantly by the for reductive elimination over b-hydrogen elimination.71 thermodynamic stability of the enolate complexes. Reductive elimination reduces the coordination number of the metal, but b-hydrogen elimination either increases σ* krel the coordination number or leaves the coordination num- R = Me 0 >600 Ph2 R = CH2Ph 0.22 >250 P Y ber unchanged. Thus, sterically hindered ligands will ac- Ar DPPBz R = CH2C(O)Ar 0.60 31 Pd R celerate reductive elimination, but will retard b-hydrogen R = CH2CF3 0.92 1.7 P R 40–100 °C Ph 50–99% R = CH2CN 1.30 1 elimination. As a result, catalysts with hindered ligands 2 R = CF 2.60 no rxn will react with better selectivity for the formation of 3 amines versus the formation of arenes. This trend is illus- Equation 11 trated by the selectivity of catalysts containing arylphos- phines of varied size, summarized in Scheme 9. Although less sensitive to electronic effects than the C–N The reaction in Equation 10 illustrates the magnitude of or C–O bond-forming reductive eliminations, the rates of the effect of sterically hindered phosphines on the rate of C–C bond-forming reductive eliminations can be large reductive elimination. The three-coordinate arylpalladium enough to control the scope of the catalytic arylation of enolates and the anions of nitriles. For example, aryl- ditolylamido complex with P(t-Bu)3 as ligand underwent palladium complexes of the anions of nitriles undergo

X reductive elimination more slowly than arylpalladium 0.001–1 mol% Pd(OAc)2 complexes of enolates because the cyano group is more Cl 0.001–1 mol% L NHR P X N 2 electron-withdrawing than an acyl group. Moreover, com- X = CH, N NaOt-Bu, DME L = Fe PCy2 plexes with two electron-withdrawing groups on the a- + HNRR' 25–70 °C, 2–16 h carbon, such as the complexes of the anion of a malonate, require hindered ligands on the metal to undergo reductive NHoctyl NHoctyl N elimination (see Scheme 10). In broad terms, the rate of N NHoctyl N N NHoctyl reductive elimination from complexes with a functional 10 ppm 100 °C 50 ppm, 90 °C 0.01%, 90 °C 50 ppm, 90 °C group on the a-carbon can be predicted from the Taft sub- 86% 93% 83% 96% 74 H H H H stituent parameter of the functional group. N N N N Bn Cy

A Ph N N N N 2 O O P o-Tol 90 °C 0.01%, 100 °C 0.01%, 100 °C 1.0%, 70 °C 0.05%, 100 °C Pd OEt EtO OEt 95% 79% 67% 91%, 95% ee P o-Tol Ph2 O H EtO N Ph O O N Ph N N (t-Bu)2 OEt H N Ph Ph N P O O Ph O FcP(t-Bu)2 Fe Pd EtO OEt 1%, 70 °C, 1%, 70 °C, 1%, 70 °C, 91% 99% 94% o-Tol O 70 °C, 1 h o-Tol OEt 95% Equation 12 Scheme 10 Our initial efforts were directed at finding a catalyst for 7 Stable Palladium Catalysts for Coupling of mild coupling of heteroaryl halides. Reactions of primary Primary Amines: Identification and Reduc- amines such as octylamine, benzylamine, isobutyl amine, tion of One Process that May Limit Turnover cyclohexylamine and tert-butylamine occurred with 2-, Numbers 3-, and 4-chloropyridine in high yields as shown in Equation 12. In many cases these reactions occurred with catalyst loadings as low as 10–50 ppm for a 1:1 ratio of The low coordination number of the palladium complexes palladium and ligand. Although we have not yet deter- with sterically hindered ligands leads to high reactivity, mined the minimum amount of catalyst necessary for re- but this low coordination number may also be an Achilles actions of pyridyl chlorides with other types of nitrogen heel. We have shown that {Pd[P(t-Bu) ](Ar)(Br)}, which 3 nucleophiles, we have shown that other primary nitrogen would be an intermediate in the coupling of aryl bromides nucleophiles, such as benzophenone hydrazone, benzo- with catalysts containing P(t-Bu) as ligand, reacts with 3 phenone imine, and benzamide react with halopyridines to pyridine and with amines to displace the phosphine ligand give the coupled products in high yields. and generate four-coordinate palladium complexes with two pyridine or amine ligands (Scheme 11). These pyri- This catalyst is also highly reactive for the coupling of dine and amine complexes are inactive for the coupling of chloroarenes. The reactions of primary amines with unac- aryl chlorides. tivated chloroarenes occurred with low loadings of a 1:1 ratio of Pd(OAc)2 and CyPF-t-Bu under mild conditions, Br Br as summarized in Equation 13. For example, the prototyp- H2NBn PhPd BnH2NPdNH2Bn ical reaction of octylamine with phenyl chloride occurred r.t., 5 min L Ph in high yield with only 50 ppm catalyst. This yield and catalyst loading correspond to nearly 19,000 turnovers, Br pyridine Br and this value is about two orders of magnitude higher PhPd Py Pd Py r.t., 5 min than the reaction of a primary amine with an aryl chloride L Ph in the presence of any other catalyst. The reactions of ali- Scheme 11 phatic amines that are branched a to nitrogen, and the reactions of benzylamine, also formed the coupled product Based on this information, we proposed that sterically in good yield with much less catalyst than has been used hindered, bidentate phosphines would generate highly ac- previously. tive catalysts for the coupling of aryl chlorides, bromides Finally, we have shown that this combination of metal and and tosylates but would resist deactivation by the replace- ligand will catalyze the reactions of primary amines with ment of the ligand by amines or basic nitrogen heteroaryl halides containing the diverse array of functional cycles. Indeed, the combination of palladium acetate and groups shown in Equation 14. These reactions were con- CyPF-t-Bu generates a spectacularly reactive catalyst for ducted under the conditions with LiN(SiMe3)2 as base to the coupling of primary nitrogen nucleophiles with aryl deprotonate protic groups.75,76 For example, the reactions 46 and heteroaryl halides. Three sets of recently reported of aryl halides bearing pendant primary or secondary 46 reactions are summarized in Equations 12– 14. alcohols, halophenols, aryl halides with enolizable

Synlett 2006, No. 9, 1283–1294 © Thieme Stuttgart · New York ACCOUNT Transition-Metal-Catalyzed Aromatic Substitution Reactions 1293

Cl 0.005–1 mol% Pd(OAc)2 NHR primary amines. Further, primary amines bind more tight- P R 0.005–1 mol% L R 2 ly to palladium(II) than do secondary amines, and the pri- NaOt-Bu, DME L = Fe PCy2 mary amines can poison catalysts with labile phosphines + HNRR' 25–70 °C, 2–16 h by displacement of the ligand. Thus, catalysts containing H octylHN NHoctyl MeO NHBn N bidentate ligands can be more reactive toward coupling of primary amines than catalysts containing monodentate Me NC ligands, and complexes of the most sterically hindered 50 ppm, 90 °C 0.01%, 100 °C 50 ppm, 80 °C 50 ppm, 80 °C 94% 98% 98% 99% version of the Josiphos ligands, CyPF-t-Bu, catalyze the H coupling of aryl bromides, chlorides and tosylates with H H NHBn N MeO N N primary nitrogen nucleophiles in high yields and with ex- Cy tremely high turnover numbers in many cases. MeO 0.05%, 100 °C 0.01%, 100 °C 0.05%, 100 °C 0.1%, 100 °C Although the origin of electronic effects can be rational- 99% 68% 99% 90% ized in many ways, reductive eliminations from complex- Equation 13 es containing less electron-donating groups bound to palladium are slower than reductive eliminations from complexes with more electron-donating groups bound to hydrogens, aryl halides with unprotected primary amides, palladium. Catalysts containing sterically hindered mono- aryl chlorides with free carboxylic acids, and chloro N- dentate phosphines, which accelerate the rate of reductive arylacetamides all reacted with primary amines in high elimination, expand the scope of coupling reactions of the yields. less electron-rich substrates by accelerating the rate of reductive elimination enough to make this desired step fast- Cl 0.05–2 mol% Pd(OAc)2 NHR R 0.05–2 mol% L R P er than competing reactions that lead to undesired side 2 LiHMDS, DME products. L = Fe PCy + HNRR' 70–100 °C 2

H H N N NHoctyl Acknowledgment HO We thank the NIH for generous support of this work and Eli Lilly, HO HO HO NHBn 0.5%, 100 °C 0.5%, 100 °C 2%, 100 °C 0.5%, 100 °C Union Carbide, DuPont, Boehringer-Ingelheim, and Merck for ge- 90% 86% 72% 84% nerous donations during early and recent stages of the project. Most important, I thank the roughly 75 members of my research group H H H N N N who have put their hearts, minds, and hands into in this work. It is O their ideas, concepts, analysis, and hard work in the laboratory that O O O N NHoctyl I report in this manuscript. H NH2 OH 0.5%, 70 °C 0.5%, 100 °C 0.05%, 100 °C 0.05%, 100 °C 92% 67% 85% 99% References and Notes Equation 14 (1) Heaney, H. Chem. Rev. 1962, 62, 81. (2) Rossi, R. A.; Bunnett, J. F. J. Org. Chem. 1973, 38, 1407. (3) Rossi, R.; de Rossi, R. H. Aromatic Substitution by the SRN1 8 Guidelines for Catalyst Selection Mechanism, ACS Monograph Series No. 178; American Chemical Society: Washington D.C., 1983. This mechanistic information provides a basis to under- (4) Hartwig, J. F. In Handbook of Organopalladium Chemistry for Organic Synthesis, Vol. 1; Negishi, E. I., Ed.; Wiley- stand the scope and the types of catalyst that are most ef- Interscience: New York, 2002, 1051. fective for reactions of the different aryl halides with (5) Hartwig, J. F. In Handbook of Organopalladium Chemistry oxygen, nitrogen, or carbon enolate nucleophiles. For ex- for Organic Synthesis, Vol. 1; Negishi, E. I., Ed.; Wiley- ample, mild reactions of unactivated aryl chlorides occur Interscience: New York, 2002, 1097. with catalysts containing phosphines with sterically hin- (6) Culkin, D. A.; Hartwig, J. F. Acc. Chem. Res. 2003, 36, 234. dered alkyl substituents, but reactions of aryl bromides (7) Kosugi, M.; Kameyama, M.; Migita, T. Chem. Lett. 1983, 927. occur with catalysts containing triarylphosphines. Fur- (8) Kosugi, M.; Kameyama, M.; Sano, H.; Migita, T. Nippon ther, different types of catalyst are most active for the re- Kagaku Kaishi 1985, 3, 547. actions of secondary amines and primary amines. The (9) For a review of our work on the reductive elimination to additional steric hindrance of secondary amines acceler- form carbon–nitrogen and carbon–oxygen bonds, see ates the rate of reductive elimination versus b-hydrogen reference 10. elimination, and many complexes with sterically hindered (10) Hartwig, J. F. Acc. Chem. Res. 1998, 31, 852. monodentate ligands catalyze the coupling of aryl halides (11) Wolfe, J. P.; Wagaw, S.; Marcoux, J.-F.; Buchwald, S. L. Acc. Chem. Res. 1998, 31, 805. with secondary amines. However, primary amines gener- (12) Muci, A. R.; Buchwald, S. L. Top. Curr. Chem. 2002, 219, ate amido complexes that undergo reductive elimination 131. slower than those generated from secondary amines. (13) Guram, A. S.; Rennels, R. A.; Buchwald, S. L. Angew. Thus, b-hydrogen elimination more often competes with Chem., Int. Ed. Engl. 1995, 34, 1348. reductive elimination during coupling of aryl halides with (14) Louie, J.; Hartwig, J. F. Tetrahedron Lett. 1995, 36, 3609.

(15) Wolfe, J. P.; Åhman, J.; Sadighi, J. P.; Singer, R. A.; (49) Rathke, M. W.; Sullivan, D. F. J. Am. Chem. Soc. 1973, 95, Buchwald, S. L. Tetrahedron Lett. 1997, 38, 6367. 3050. (16) Mann, G.; Hartwig, J. F.; Driver, M. S.; Fernandez-Rivas, C. (50) Jørgensen, M.; Liu, X.; Wolkowski, J. P.; Hartwig, J. F. J. J. Am. Chem. Soc. 1998, 120, 827. Am. Chem. Soc. 2002, 124, 12557. (17) Lee, S.; Jorgensen, M.; Hartwig, J. F. Org. Lett. 2001, 3, (51) Hama, T.; Liu, X.; Culkin, D.; Hartwig, J. F. J. Am. Chem. 2729. Soc. 2003, 125, 11176. (18) Huang, X. H.; Buchwald, S. L. Org. Lett. 2001, 3, 3417. (52) Shaughnessy, K. H.; Hamann, B. C.; Hartwig, J. F. J. Org. (19) Hartwig, J. F. Angew. Chem., Int. Ed. Engl. 1998, 37, 2090. Chem. 1998, 63, 6546. (20) Wagaw, S.; Yang, B. H.; Buchwald, S. L. J. Am. Chem. Soc. (53) Liu, X.; Hartwig, J. F. J. Am. Chem. Soc. 2004, 126, 5182. 1998, 120, 6621. (54) Our original work on the catalytic process led to the (21) Wagaw, S.; Yang, B.; Buchwald, S. L. J. Am. Chem. Soc. conclusion that the immediate product of the C–N bond- 1999, 121, 10251. forming step [Pd(BINAP)(NArRR')], reacts with free ligand (22) Haddad, N.; Baron, J. Tetrahedron Lett. 2002, 43, 2171. to regenerate [Pd(BINAP)2] faster than the amine dissociates (23) Mann, G.; Hartwig, J. F. J. Org. Chem. 1997, 62, 5413. to generate [Pd(BINAP)] and that [Pd(BINAP)2] lies directly (24) Mann, G.; Hartwig, J. F. Tetrahedron Lett. 1997, 38, 8005. on the catalytic cycle. We recently remeasured kinetic data (25) Mann, G.; Hartwig, J. F. J. Am. Chem. Soc. 1996, 118, on the order of the catalytic process in bromoarene and 13109. ligand and our corrected data imply that (26) Palucki, M.; Wolfe, J. P.; Buchwald, S. L. J. Am. Chem. Soc. [Pd(BINAP)(NArRR')] dissociates the amine to form 1997, 119, 3395. [Pd(BINAP)] faster than it reacts with the free ligand and (27) Palucki, M.; Wolfe, J. P.; Buchwald, S. L. J. Am. Chem. Soc. that [Pd(BINAP)2] lies off the cycle as shown in Scheme 4. 1996, 118, 10333. These data are presented in reference 62. (28) Mann, G.; Incarvito, C.; Rheingold, A. L.; Hartwig, J. F. J. (55) Fauvarque, J.-F.; Pflüger, F. J. Organomet. Chem. 1981, Am. Chem. Soc. 1999, 121, 3224. 208, 419. (29) Shelby, Q.; Kataoka, N.; Mann, G.; Hartwig, J. F. J. Am. (56) Amatore, C.; Pfluger, F. Organometallics 1990, 9, 2276. Chem. Soc. 2000, 122, 10718. (57) Hartwig, J. F.; Paul, F. J. Am. Chem. Soc. 1995, 117, 5373. (30) Kataoka, N.; Shelby, Q.; Stambuli, J. P.; Hartwig, J. F. J. (58) Barios-Landeros, F.; Hartwig, J. F. J. Am. Chem. Soc. 2005, Org. Chem. 2002, 67, 5553. 127, 6944. (31) Aranyos, A.; Old, D. W.; Kiyomori, A.; Wolfe, J. P.; (59) Alcazar-Roman, L. M.; Hartwig, J. F.; Rheingold, A. L.; Sadighi, J. P.; Buchwald, S. L. J. Am. Chem. Soc. 1999, 121, Liable-Sands, L. M.; Guzei, I. A. J. Am. Chem. Soc. 2000, 4369. 122, 4618. (32) Torraca, K.; Kuwabe, S.; Buchwald, S. J. Am. Chem. Soc. (60) Singh, U. K.; Strieter, E. R.; Blackmond, D. G.; Buchwald, 2000, 122, 12907. S. L. J. Am. Chem. Soc. 2002, 124, 14104. (33) Kuwabe, S.; Torraca, K. E.; Buchwald, S. L. J. Am. Chem. (61) Shekhar, S.; Ryberg, P.; Hartwig, J. F. Org. Lett. 2006, 8, Soc. 2001, 123, 12202. 851. (34) Parrish, C. A.; Buchwald, S. L. J. Org. Chem. 2001, 66, (62) Shekhar, S.; Ryberg, P.; Hartwig, J. F.; Mathew, J. S.; 2498. Blackmond, D. G.; Strieter, E. R.; Buchwald, S. L. J. Am. (35) Torraca, K. E.; Huang, X. H.; Parrish, C. A.; Buchwald, S. Chem. Soc. 2006, 128, 3584. L. J. Am. Chem. Soc. 2001, 123, 10770. (63) Stambuli, J. P.; Bühl, M.; Hartwig, J. F. J. Am. Chem. Soc. (36) Hamann, B. C.; Hartwig, J. F. J. Am. Chem. Soc. 1997, 119, 2002, 124, 9346. 12382. (64) Stambuli, J. P.; Incarvito, C. D.; Buhl, M.; Hartwig, J. F. J. (37) Palucki, M.; Buchwald, S. L. J. Am. Chem. Soc. 1997, 119, Am. Chem. Soc. 2004, 126, 1184. 11108. (65) Barrios-Landeros, F.; Hartwig, J. F. J. Am. Chem. Soc. 2005, (38) Ehrentraut, A.; Zapf, A.; Beller, M. Adv. Synth. Catal. 2002, 127, 6944. 344, 209. (66) Roy, A. H.; Hartwig, J. F. Organometallics 2004, 23, 194. (39) Viciu, M. S.; Germaneau, R. F.; Nolan, S. P. Org. Lett. 2002, (67) Driver, M. S.; Hartwig, J. F. J. Am. Chem. Soc. 1995, 117, 4, 4053. 4708. (40) Fox, J. M.; Huang, X. H.; Chieffi, A.; Buchwald, S. L. J. Am. (68) Driver, M. S.; Hartwig, J. F. J. Am. Chem. Soc. 1997, 119, Chem. Soc. 2000, 122, 1360. 8232. (41) Kawatsura, M.; Hartwig, J. F. J. Am. Chem. Soc. 1999, 121, (69) Hartwig, J. F. J. Am. Chem. Soc. 1996, 118, 7010. 1473. (70) Zhao, J.; Hesslink, H.; Hartwig, J. F. J. Am. Chem. Soc. (42) Driver, M. S.; Hartwig, J. F. J. Am. Chem. Soc. 1996, 118, 2001, 123, 7220. 7217. (71) Hartwig, J. F.; Richards, S.; Barañano, D.; Paul, F. J. Am. (43) Nishiyama, M.; Yamamoto, T.; Koie, Y. Tetrahedron Lett. Chem. Soc. 1996, 118, 3626. 1998, 39, 617. (72) Yamashita, M.; Hartwig, J. F. J. Am. Chem. Soc. 2004, 126, (44) Hartwig, J. F.; Kawatsura, M.; Hauck, S. I.; Shaughnessy, K. 5344. H.; Alcazar-Roman, L. M. J. Org. Chem. 1999, 64, 5575. (73) Culkin, D. A.; Hartwig, J. F. J. Am. Chem. Soc. 2001, 123, (45) Stambuli, J. P.; Kuwano, R.; Hartwig, J. F. Angew. Chem. 5816. Int. Ed. 2002, 41, 4746. (74) Culkin, D. A.; Hartwig, J. F. Organometallics 2004, 23, (46) Shen, Q.; Shekhar, S.; Stambuli, J. P.; Hartwig, J. F. Angew. 3398. Chem. Int. Ed. 2004, 44, 1371. (75) Huang, X. H.; Anderson, K. W.; Zim, D.; Jiang, L.; Klapars, (47) Roy, A. H.; Hartwig, J. F. J. Am. Chem. Soc. 2003, 125, A.; Buchwald, S. L. J. Am. Chem. Soc. 2003, 125, 6653. 8704. (76) Harris, M. C.; Huang, X.; Buchwald, S. L. Org. Lett. 2002, (48) Rathke, M. W.; Lindert, A. J. Am. Chem. Soc. 1971, 93, 4, 288. 2318.