The syn/anti-Dichotomy in the Palladium-Catalyzed Addition of Nucleophiles to Alkenes

In this review the stereochemistry of palladium-catalyzed addition of nucleophiles to alkenes is discussed, and examples of these reactions in organic synthesis are given. Most of the reactions discussed involve oxygen and nitrogen nucleophiles; the Wacker oxidation of ethylene has been reviewed in detail. An anti-hydroxypalladation in the Wacker oxidation has strong support from both experimental and computational studies. From the reviewed material it is clear that anti-addition of oxygen and nitrogen nucleophiles is strongly favored in intermolecular addition to olefin–palladium complexes even if the nucleophile is coordinated to the metal. On the other hand, syn-addition is common in the case of intramolecular oxy- and amidopalladation as a result of the initial coordination of the internal nucleophile to the metal.


Introduction
Electrophilic addition to olefins is one of the fundamental reactions in organic chemistry.T hus, bromine or hypobromous acid are readily added across an electron-rich olefinicd ouble bond (1)i naprocess that is initiated by an electrophilica ttack to generate the corresponding bromonium ion 2,w hich is then opened by Br À (in the case of Br 2 )o rH 2 O( when HOBr is employed) from the opposite side, to afford the anti-addition product (Scheme 1). [1,2] Electrophilic metal cations M n + ,s uch as Hg 2 + and Tl 3 + ,f ollow as imilarp attern in both inter-and intramolecular reactions. [1,3] By contrast, owing to the nature of the electron-rich C=Cb onds, nucleophilic additions are rare and typicallyl imited to intramolecular additions of an alkoxide moiety,g enerated from as uitable polycyclic alkenol by deprotonation with NaH or tBuOK. [4] Transition metals are known to form h 2 -complexes 3 that can be attacked by nucleophiles to effect af ormal nucleophilic addition upon the removal of the metal. [5] However,t he transition-metal-promotedp rocess is more complicated, since the nucleophile can, ap riori, attack the h 2 -species either from the opposite side (3)i na nalogy to bromonium ions 2 and their congeners, or could first coordinate to the metal and then be delivered in a syn-fashion to the originalalkene (4). [5] The first h 2 -olefin metal complex 5 was prepared by Zeise (Scheme 2) via coordinationo fe thylene (generated by an in situ dehydrogenation of boilinge thanol) to K 2 PtCl 4 . [6] However, this complex was long considered as ar arity andi tt ook more than 100 years before the nature of the p bond between an alkene and am etal wasu nderstood and its potentialh ad been realized.
The present review is concerned with the stereochemistry of nucleophilic additions to h 2 -alkene metal complexes of palladium (analogous to 5)a nd application of these reactions in synthetic organic chemistry.The experimental and theoretical findings, accumulatedo ver the years, are submitted to at horough mechanistic analysisi nl ight of the mostr ecent results, which are not included in previousr eviews. The new resultsh ave allowed us to draw conclusions that were not explicitly expressed in previousreviews. [7] generated in the previousstep [Eq. (2)].The latter reaction produces CuCl, which is then reoxidized to CuCl 2 by molecular oxygen, consuming the two equivalents of HCl generated in the first step [Eq. (3)].T he reactioni st hus catalytic in both Pd II and Cu II ,w ith molecular oxygen serving as the terminal, stoichiometric oxidant. Note that directr eoxidation of Pd 0 to Pd II by molecular oxygen in water under Wacker conditions is too slow and would not preventa ggregation and formation of metallicpalladium.
After the original publication on the Wacker process, this reaction becamet he subjecto fn umerousm echanistic studies. [7a] It was proposed that the intermediate h 2 -complex 7 reacts with water to produce the 2-hydroxyethylpalladium complex 8,w hich would undergo a b-H elimination to generate vinyl alcohol 10 (via 9a), tautomerization of which would produce acetaldehyde 11 (Scheme 4). However,d euterium labeling studies clearly demonstrated that the final stages of the cascade do not adopt this route:t he h 2 -complex 9a,p rimarily arising from 8 by the expected b-H elimination, does not dissociate to produce 10,b ut rather undergoes an insertion reaction to generate complex 12 (via rotamer 9b), which now uses the OÀH( rather than CÀH) for the final b-H elimination (12 a or 12 b)t op roduce acetaldehyde (11). [9] An alternative pathway,w here al one-pair on oxygen ejects Pd 0 as al eaving group (12 b), was also considered, [10] which was later supported by theoretical calculations. [11] The latter mechanism has been inferred from isotopic labeling as follows (Scheme 5): If the reactionp roceeded through enol 10,t he molecule would abstract ap roton from the envi-Abstracti nC zech: Vt omto přehlednØm člµnku je diskutovµna stereochemie palladiem katalyzovanµ adice nukleofilů na alkeny aj sou zde uvedeny příklady aplikací vo rganickØ syntØze. Většina diskutovaných reakcí zahrnuje kyslíkatØ ad usíkatØ nukleofily.Z vlµštníp ozornost je věnovµna oxidaci ethylenu, kterou vyvinulaf irma Wacker.Vtomoto případě experimentµlní iv ý početní studie jasně ukazují na anti-hydroxypalladaci jako dominantní mechanisms. Zd ostupnØho materialu vyplývµ, že antimechanismus je výrazně preferovµn vp ř ípadě intermolekulµrních adicí kyslíkatých ad usíkatých nukleofilů na komplexy olefin-palladium,at oitehdy,j e-li nukleofil koordinovµn ke kovu. Naopak, syn-adice je běžnµ vp ř ípadě intramolekulµrní oxy-aa midopalladace jako výsledek prvotní koordinacei nterního nukleofilu ke kovu. ronment [note that there is free HCl available, according to Equation (1). However,t he experiment with perdeuteriated ethylene [Eq. (4)] showed that all the label wasr etained, which is consistent with the mechanism depicted in Scheme 4, namely with the isomerization 9a ! 9b ! 12.T he isotope effect [compareE qs. (1) and (4) in Scheme 5] was found to be marginal. [9] In ac omplementary experiment, the specifically dideuteriated ethylenes 13 and 14 (Scheme 6) were found to exhibit an internal competitive isotope effect (H vs Ds hift), demonstrating that the rate-limiting step must occur before the b-H elimination and formation of acetaldehyde. [12,13] It was further argued that the rate-limiting step is the hydroxypalladation, and the results are better explained by a syn-migration since an anti-hydroxypalladation would seem to require ar ate constant larger than diffusion control. [9,12] The proposed syn-migration of an oxygen nucleophile has been experimentally observed in as toichiometrice xperiment with aPt-olefin complex (Scheme 7). In this experiment the extremely electron-deficient tetrafluoroethylene (15)w as treated with the Pt-OMe complex 16.T he formation of intermediate 17 and product 18,a sw ell as the kinetics, were monitored by NMR spectroscopy in [D 8 ]THF with added CD 3 OD. [14] Here, an externala ttack by CD 3 OD (whichs hould proceed with anti stereochemistry due to the steric hindrance from the Pt side) has not been observed, nor was the exchange of OCH 3 with OCD 3 , so that the product 18 can only arise by the syn-migration of OCH 3 from Pt. This may seem to support the syn-migration pathway in the Wacker process. However,i tc an be argued that the starting olefin 15 in this experiment is rather special, so that generalization of these findings, in particular to the Wacker oxidation of the electron-rich ethylene, cannot be made directly.
In order to elucidate the stereochemistry of hydroxypalladation, the trans-dideuteriated ethylene 19 wase mployed as am odel compound (Scheme 8). [10,15] Under catalytic hydroxypalladation conditions, using am ixture of PdCl 2 ,L iCl, and CuCl 2 ,t he latter alkene was expected to generate the h 2 -complex 20,w here the water molecule would have to be cis-coordinated toward the ethylene ligand (otherwise the syn-transfer would not be possible at all). The anti-attack by an external water molecule on 20 should generate complex 21,w hich in the presence of chloridei ons should produce chlorohydrin 22 as ar esult of the S N 2d isplacemento fp alladium. Treatment of 22 with ab ase would then produce epoxide 23.T he latter epoxide was found to be cis-configured (as shown), which is consistent with the pathway involvingt hree inversions of configuration, that is, in each step starting with complex 20.S ince the replacement of palladium by Cl À was known from the previous work [16] to proceed with inversion of configurationa tc arbon, and the stereochemistry of the chlorohydrin transformation into the corresponding epoxide is atextbookexample of inversion, the key hydration 20 ! 21 must also occur with inversion, that is, via an externala ttack as shown. This study thus provides evidence for the anti-addition mechanism for the key hydroxypalladation step.
An extended theoretical study [17] then reconciled the fact that depending on the nature of the nucleophile the nucleophilic attack may occur either in a syn or anti fashion (Scheme 9). Essentially,t wo types of nucleophiles can be discerned:N u A ,w hich can, ap riori, coordinate the metal but prefer the external anti-addition to the olefinic ligand (pathway a), and Nu B ,w hich prefer an intramolecular,that is, syn-transfer (pathway b).
Analysis of the orbitali nteractions in four model systems, namely with OH À ,F À ,H À ,a nd CH 3 À as representative nucleophiles coordinated to Pd, revealed the following (Figure 1): [17b] the HOMO orbitals of the Scheme5.Deuterium labeling in Wacker oxidation.
Scheme7.Experimentally observed syn-migration of MeO from platinum.
Scheme8.Stereochemistry of hydroxypalladation in the presence of chloride ions. first two complexes lie rather low,s ot hat their interaction with the LUMO of the (coordinated) olefin will be weak and result in little stabilization.B yc ontrast, the two latter complexes have high energy HOMO orbitals for the PdÀNu bond (Nu = H À and CH 3 À ), so that the interaction between the HOMO orbital of these complexes and the LUMO of the alkene shouldb e stronger and lead to ac onsiderable loweringo ft he energy of the system.H ence, it can be anticipated that the reactivity of H À and CH 3 À will be controlled mainly by orbital interactions, so that the transfer from the metal should be preferred. On the other hand, the orbitale ffects in the case of OH À and F À are likely to be weak, so that the reaction should be dominated by ionic effects and thus proceed via anti-addition. [17b] Furthermore, the reactions controlled by ionic interaction (favoring the anti-pathway) can be expected to obey the Markovnikovr ulel ike any other electrophilica ddition and afford products,w here the incoming nucleophile is planted on the carbon that can better stabilizeapartial positive charge (24 in Scheme 10). On the other hand, in the orbital-controlled reactions (favoring the syn-pathway) the nucleophile should add to the less-substituted carbon (25), as the LUMO orbital at that carbon should be larger. [17] Ap ossible supramolecular interaction, involving severalm olecules of water and hydrogen bonding to the chloride bound to Pd, has also been considered for the Wacker oxidation ( Figure 2). [18] It has been argued that the chloride concentration in the experimentss hown in Scheme 8i sh igher (3 m)t han in the Wacker process and in the kinetic experiments [9] ( 1 m)a nd that the stereochemistry may not be the same at the different concentrations of Cl À . [7,19,20] However,t he chemistry of 1,4-difunctionalization of dienes,c atalyzed by Pd(OAc) 2 ,s hows that while under chloride-free conditions, the acetate group is delivered preferentially to the intermediate allyl group from Pd (i.e.,i nt he syn-fashion), addition of only catalytic amountso fL iCl (4 equiv per Pd) shuts down this reaction pathway.T he chloride is replacing coordinated acetate on Pd, and in this way acetate can only be delivered in an anti-fashion by an intermolecular reaction. [21] Further increase of the chloride concentration does not alter the latters tereochemistry and only increases the proportion of the product resulting from the Cl À attack on the h 3 -complex (rather than AcO À attack). This is in direct analogy to the formation of chlorohydrin 22 in the Wacker process, lending additional support for the argument that the stereochemistry should be the same at high and moderate concentrationso f chloridei ons.
Kinetic studies have demonstrated that the Wacker oxidation is actually inhibited by an increasing concentration of chloride ions [Eqs. (5)-(7) in Scheme 11]. [7,17,19] In fact, at 1 m concentration of Cl À and CuCl 2 (industrial conditions for Wacker oxidation), acetaldehyde is the predominantly formed product (with small amounts of the corresponding chlorohydrin). By contrast, at 3 m concentration of Cl À (conditions used in the mechanistic study shown in Scheme 8 [15] ), chlorohydrinb ecomes the predominant product (Scheme 12). It is highly unlikely that ac hange of the chloride ion concentrationf rom 1 m to 3 m would change the stereochemistry of the hydroxypalladation of ethylene.
Finally,s trong support for the anti-addition of AcO À and Pd II under the chloride-free conditions has been obtained from the acetoxypalladation of deuteriated 3,3-dimethyl-1-butene 26 (Scheme 13). [22] Here, the isomeric products 30 and 31 were shown to arise from predominant anti-attack of AcO À on the    , followed by the stereospecific b-hydrogen elimination from the respectivei ntermediates 28 and 29 that occurs solely in a syn-fashion. Some loss of the stereochemical integrity observed for this process was attributed to ap artial isomerization of the Markovnikov product (confirmed by ac ontrol experiment), so that the (E)and (Z)-vinyl acetates 30 a and 30 b were obtained as a3 .5:1 mixture. The anti-Markovnikov product was obtained as a9 :1 mixture of (E)-isomers 31 a and 31 b.H ere, the control experiment with non-deuteriated 3,3-dimethyl-1-butene showedt hat the Markovnikov product was ap ure (E)-isomer,a pparently arising from ac onformation of the non-deuteriated congener of 29,w here the bulky tBu and AcO groups avoid the gauche interaction. The 9:1r atio of 31 a and 31 b thus suggests that the initial acetoxypalladation proceeds mainly (but not solely) via the antimechanism. Complementary results were obtained with the (E)-isomer of 26. [22] In ac omputational study, [11] the energy of various intermediates and transition states that can be considered for the hydration pathway was assessed (Scheme 14). For the scenario with low concentration of Cl À ,t he starting tetrachloropalladate 32 is first converted into the h 2 -complex 6,w hich can be attacked by water either at Pd or at the alkene ligand. The energy of the corresponding transition states 33 and 34 has been calculated to be almost identical but the Pd-hydrated product 35,a rising from the former TS ¼ 6 ,i sb y1 0.7 kcal mol À1 lower in energy than 36,a rising from the latter TS ¼ 6 .D eprotonation of 35 to generate 37,w hich in principle could transfer the hydroxy group to the alkene ligand, would proceedt hrough the transition state 38 but this species is rather high in energy (33.4 kcal mol À1 ), well above the experimentally overall observed value (22.4 kcal mol À1 ).
However,a na lternative pathway,n amely the water-assisted deprotonation of 35 with concomitant syn-migration, eventually leading to 41,w ould proceed through the transition state 39 that is half-way down on the energy scale compared to 38, indicating that this pathway would be favored if the syn addition mechanism operated. Conversely,t he anti-delivery of water (34)g enerates intermediate 36.E ither of the transition states (39 or 40)w ould give rise to 41 by water-assisted deprotonation. Dissociation of the PdÀOb ond in 41 would generate complex 43 ready for the expected b-H elimination. The calculations further found the energy of the corresponding transition state 42 (including the agostic Pd-H interaction) to be 23.2 kcal mol À1 ,w hich makes this species highest in energy Scheme11. Chloride effect:r ate inhibition.
Scheme12. Chloride effect:c hange of mechanism and product distribution.
Scheme14. Energies of various intermediates in the proposed syn-mechanism of the Wacker oxidation in the absence of CuCl 2 . in the whole cascade (via 39/ 40), [11] so that this step can be regarded as rate-limiting.
However,t here is ap roblem with this calculation since the transition state for the antiattack is calculated on the negatively charged species 6.T he kinetics of the Wacker process shows that two chlorides are displaced, so an anti-attack would occur on the neutral species PdCl 2 (OH 2 )(CH 2 =CH 2 ). It can be expected that the latter neutral species reacts much faster than the negatively charged species 6 in external anti-attack by water.T his has also been confirmed later in other calculations (vide infra).
Subsequentc alculations [23] suggested that the anti-addition mechanism is strongly favored over the syn pathway;t his was disputed in as ubsequent article [24] and the quality of the calculation was questioned.
Calculationsi ncluding CuCl 2 gave as lightly different picture, [11] as shown in Scheme 15 (the structures here bear the same numbersa si nS cheme 14 but with added "Cu" so that ad irect comparison can be made). Here, the initial PdCl 2 /CuCl 2 complex 32-Cu first coordinates ethylene to produce 6-Cu (upon al osso fC l À ), which then can be attacked by water either at Pd or at the carbon via transition states 33-Cu and 34-Cu,r espectively,o fw hich the latter is lower in energy by 4.0 kcal mol À1 ,i ndicating that the anti-pathwayi sp referred. Furthermore, the energy of the TS ¼ 6 33-Cu is 26.0 kcal mol À1 ,w hichi ss lightly highert han the experimentally observed value (22.4 kcal mol À1 ). The two transition states would generate the hydrated species 35-Cu and 36-Cu that are almost equal in energy but the subsequent gradualc onversion into 41-Cu should preferably proceed through TS ¼ 6 40-Cu that is by 4.4 kcal mol À1 lower in energy than its congener 39-Cu.T he end-game from 41-Cu via 42-Cu and 43-Cu has been calculated to require lower energy in each step than the experimental value. Hence, the highest energy in this scenario (not exceeding the experimental value) is that of the transition state 34-Cu,s ot hat the conversion of 6-Cu into 36-Cu can be regarded as the rate-limiting step (RLS) and the reactions hould proceed predominantly via an anti-addition. [11] Again, one can argue that externala ttack on an (ethylen)palladium complex, where water has coordinated to Pd to remove negative charge (vide supra), should react faster than the negatively charged complex 6-Cu (34-Cu).
Recent computational studies, whichi ncluded aqueous medium, namely ac ubic box containing up to 26 molecules of water as am odel that is more closely related to the "real" experimental conditions, arrived at the conclusion that anti-hydroxypalladation is the favoreds tereochemical pathway (Scheme 16): [25] First, the initial coordination of tetrachloropalladate to ethylene (44)i sk nown to produce the h 2 -complex 6, which was taken as the starting point. The latterc omplex should then undergo al igand exchange if syn-migration of the nucleophile from Pd is to be allowed. In view of the geometrical restriction, previous study [11] only considered the cis-complex 35.H owever,a st he new study shows, the calculated activation barrier for its formation from 6 is 35 kcal mol À1 ,w hich is more than twice as high as that for the pathway leadingt oi ts trans-isomer 45 (14 kcal mol À1 ). Since the experimental value for the whole Wacker process has been found to be only 22.4 kcal mol À1 ,t he formation of 35 appearsu nlikely.N ote that the conversion of 6 into 45 will be assisted by the trans-effect (absent in the formation of 35), which should considerably lower the activation energy,c onsistentw ith the calculations. Furthermore, the activation energy required for the syn-migration to generate complex 47 from 35 has been calculated to be 60 kcal mol À1 ,f ar above the experimental value. By contrast, the activation barrier for the anti-attackb yw ater on the transcomplex 45 to produce 46 was calculated to be only 19 kcal mol À1 .H ence, it can be clearly seen that none of the activation barriers in the anti-pathway (6 ! 45 ! 46)exceeds the experimentally established value [11] for the whole process (22.4 kcal mol À1 ), also indicating that the rate-limiting step should occur Scheme15. Energies of various intermediates in the proposed syn-mechanism of the Wacker oxidation in the presence of CuCl 2 .
Chem. Eur.J.2015, 21,36-56 www.chemeurj.org after those initial steps, which is consistent with the previous findings. [11] The mechanismi nS cheme 16 is also consistent with the rate expression of the Wacker reaction [Eq. (7)],w here the rate is inverselyd ependento n [H + ]a nd the squareo f[ Cl À ]. These advanced calculations thus allow to conclude that the anti-hydroxypalladation should be the preferred pathway. [25] Another theoretical study,i nc ombination with experimental results (Scheme 17), provides additional support fort he anti-hydroxypalladation pathway. [26] In the latter study,t he equilibria in the system of ethylene, tetracholoropalladate, and water were investigated. Starting with the tetrachloropalladate (32), there are two initial reactions to be considered: aligand exchange with ethylene, generating h 2 -complex 6,a nd partial hydrolysis, producing complex 48.
Calculations suggest that the equilibrium should favor 48,a st he energy of its formation is lower than that of 6 (by 5.1 kcal mol À1 ). The experimentally established difference was found to be smaller (0.5 kcal mol À1 ), indicatingt hat both species should be considered for further reactions. Ligand exchange with water in the case of 6 would produce either the trans-isomer 45 or its cis-counterpart 35.A sd iscussed in the previousp aragraph, [25] conversion of 6 into 45 should be associated with ab arrier of 14.4 kcal mol À1 ; the reversed process would require 17.0 kcal mol À1 ,s ot hat this equilibrium should be shifted toward 45.F ormation of the ciscomplex 35 (from 6)w ould require 22.6 kcal mol À1 ,a ccording to these calculations, whereas the reversed reaction can proceed with ab arrier of only 14.4 kcal mol À1 ,s ot hat the equilibrium should favor the starting complex 6.I no ther words, the displacemento fC l À in 6 by water should preferentially produce the trans-complex 45.T he other possible mechanism would be the ligand exchange startingw ith the aqua-complex 48.I ts reaction with ethylene, generating the trans-complex 45 was found to be associated with the activation energy of 23.8 and 24.5 kcal mol À1 for the forward and reversed process, respectively.T he latter difference is sufficiently small to allow the existence of both species in the equilibrium mixture. Ac om-pletely different scenario wasf ound for the conversion of 48 into the cis-complex 35.H ere, the forward process would require 19.4 kcal mol À1 ,w hereas the reversed reactionw ould be associated with merely a6 .1 kcal mol À1 barrier, showing that the equilibrium should be heavily shiftedt oward 48.T hese finding are thus consistentw ith the previously suggested dominance of the trans-complex 45 over its cisisomer 35.S ince the barrierf or the anti-attacko n45 was previously calculated [25a] to be 19 kcal mol À1 , whereas the syn-migration from 35 would require [25a] 60 kcal mol À1 ,the syn-pathway can be ruled out. [26] When extended beyond the original conversion of ethylene to acetaldehyde, the Wacker oxidation has served, with variousm odifications, as as tandard method for oxidation of terminal olefins to produce methyl ketones [27,28] (e.g., 49 ! 50 in Scheme 18); [29] this process is referred to as the Wacker-Tsuji oxidation. [27] However,a lteration of the regioselectivity in favor of the corre-Scheme16. Molecular dynamics calculations including ac ubic box of 26 molecules of H 2 O, which simulates the "real" reaction medium and low Cl À concentration.
Scheme18. Recent modifications of the Wacker oxidation of terminal olefins and alteration of its regioselectivity.
Chem. Eur.J.2015, 21,36-56 www.chemeurj.org sponding aldehydes 52 has recently been achieved simply by using tBuOH rather than water.T he reaction presumably proceeds via the vinyl ether 51,r esulting from the anti-Markovnikov attack of the bulky nucleophile at the sterically less hindered terminal carbon. [30] 3. Amination of Olefins Relatedt oW acker oxidation, where the stereochemistry of the initial hydroxypalladationw as discussed in detail in the previous chapter,i st he amination of olefins. However,i nc ontrast to water and alcohols, amines are not normally considered for analogousP d-catalyzed reaction, as they are easily oxidized. Nevertheless, as toichiometricP d-mediated amination of (E)-2butene (53)w ith dimethyl amine has been successfully investigated at low temperature (Scheme 19). [31] The initially formed aminopalladation product 54 was reduced in situ with LiAlD 4 (with retention of configuration) to afford the deuteriated amine 55.
[31b] The relative configurationo ft he latter derivative was established in two steps, involving N-oxidation (55 ! 56) followed by Cope elimination, which is known to proceed with a syn-elimination mechanism. The product analysis (57)(58)(59) was consistentw ith an initial anti-addition of Pd II and Me 2 NH across the C=Cb ond, [31b] in analogy to the mechanism discussed fort he Wacker oxidation. Complementary experiments with (Z)-2-butene led to the same conclusions. [31,32] Note that before the amine attacks the coordinated olefin, two amine molecules coordinate to palladium. A syn-migration of the amine from Pd to the coordinated olefin should be precluded, as this process apparently would be much higher in energy (Figure 1) than the intermolecular anti-attack by Me 2 NH, generating 54.T his is in contrastt ot he Pd-OAc h 3 -species, where the syndelivery via ac yclic transition state is allowed. [33] However,a nalogous syn delivery in the case of the corresponding h 2 -complexes is disfavored (Scheme 13). [22]

Intramolecular Oxypalladation
The stereocontrolled intramolecular nucleophilic attack on am etal-olefin complex (asi n60), is an important reaction in synthetic organic chemistry, [34] and is analogous to halolactonization, haloetherification, and related reactions employing S, Se, Hg,Tl, Au, and other electrophiles [Scheme 20, Eq. (8)]. [1,35] When metals,s uch as Pd [1,3,5] or Hg, [1,3] are employed as the electrophilic triggerso ft he reaction, the initially generated organometallicp roduct 61 can be utilized in as ubsequentr eactiont hat would allow the construction of an ew CÀCb ond from the CÀMb ond. Theo verall result would then be the formation of aC -X and CÀCb ond, where Xi si ntroduced as an ucleophile,a nd the new C-substituentf ormally as an electrophile (62). [1] Aside from this antimechanism [Eq. (8)],t he syn-addition 63 ! 64 [Eq. (9)] may also operate (see also Scheme 1), due to the initial coordination (60 ! 63), known for instance from the vanadium-catalyzed epoxidation. [1c] Examples of both mechanisms and the effects favoring one or the otherw ill be discussed and analyzed in this and subsequent sections.

Intramolecular Oxypalladation FollowedbyC arbonylation
The first intramolecular oxypalladation was carried out in conjunction with carbonylation (Scheme 21). [36] Here, the reaction of alkenol 65 with CO and MeOH,c atalyzed by Pd II ,c ommencedw itht he h 2 -coordination, followed by an anti-attack by the neighboring hydroxyl group. Of the two facial stereoisomers, one is consumed faster (66)a nd generates the tetrahydropyran derivative 67 with high stereocontrol exercised by the original chiralc enter (note the "equatorial" methyl in 66). Coordination of carbon monoxide( 68), followed by migratory Scheme19. Stereochemistry of the Pd II -catalyzed amination of 2-butene.
Scheme20. Neighboring group effects in the metal-catalyzed functionalization of olefins.
Scheme21. Intramolecular hydroxypalladation controlled by ar esiding chiral center followed by carbonylation. www.chemeurj.org insertiont hen generated complex 69,w hich on reaction with methanolp roduced the desired methyl ester 70 and Pd 0 , whose reoxidation with Cu II completed the catalytic cycle.
Several other examples followed shortly,a ll featuring the initial anti-addition of the electrophilic Pd II and the neighboring OH group across the C=Cb ond, [37][38][39] such as the cyclization of 71 and 73 (Scheme 22), [39] and 76 (Scheme 23), followed by carbonylation, as detailed in the previousparagraph. [3a] All those reactions (Schemes 21-23) occurred with alkenols, whose structures allow 5-exo-o r6 -exo-cyclizations. If the only option is a4 -exo process, as in the case of the homoallylic alcohol 78,t he reaction has been found to take ad ifferent course (Scheme 24). [40] Here, the C=Cb ond coordination to Pd II (as described in Scheme 21) becomes unproductive, as cyclization to produce the corresponding oxetane 79 would be too high in energy.I nstead, the reaction proceeds through ad ifferent channel, namely that involving the initial coordination of CO to Pd, followed by ar eactionw ith the OH group and additional coordination to the C=Cb ond (80). These eventsa re followed by reductivee limination to generate lactone 81 with Pd chelated to the exocyclic carbon in an h 1 -fashion( which formally corresponds to a synaddition across the C=Cb ond). The cascade is then completed by the second carbonylation to produce 82 (Scheme 24). The trans-configured alcohol 83 reacted in the same way to produce the diastereoisomeric lactone 84. The reaction conditions for theset ransformations are noteworthy:a side from the standard reoxidation of the resultingP d 0 by Cu II ,p ropene epoxide was employedt oc onsume the Brønsted acid generated by the reaction, whereas trimethyl orthoacetatew as added to secure anhydrous conditions. [40] Other examples of this reaction course,where the cyclization involving intramolecular oxypalladation is precluded by structural restrictions in the homoallylic arrangement, have been reported but without addressing the stereochemistry issues. [41]

Intramolecular Oxypalladation Followedbyb -Hydride Elimination
The stereochemistry of the Pd II -catalyzed ring closure of alkenols has been investigated with the aid of the stereospecifically deuteriated substrate 85,u sing p-benzoquinone (p-BQ) as the terminal oxidant( Scheme 25). [42] Sincei ntermediates in ac atalytic reaction are difficult to isolate, the steric course was inferred from the products of the subsequent b-H elimination, which is known to occur as a syn-process.W ith the weakly coordinating BF 4 À anion at Pd II ,t he free phenolic hydroxyl group tends to coordinate to the metal [as in Eq. (9), Scheme 20], which leads to a p-olefin complex where Pd is bound to the top face of cyclohexene. A syn-oxypalladation then produces 86,w hich on subsequent b-hydride elimination removed the cis-positioned deuterium. By contrast, coordination to the neighboring hydroxyl is precluded in the case the strongly co-Scheme22. Carbonylative cyclization:t he effect of alkene geometry on the regioselectivity.
Scheme25. Switching between syn-and anti-mechanism in the intramolecular hydroxypalladation as af unction of the anion coordinated to Pd. ordinating Cl À at Pd II (especially when additional LiCl is present) and the reaction thus proceeded with anti-stereochemistry to give 87,a sr evealed by the b-H elimination of the cis-disposed proton. [33,42,43] The latter study used phenolic hydroxyl as an eighboring group, whose coordination capability can differ from that of an ordinarya lcohol due to the difference in the pK a .T herefore, two diastereoisomerically deuteriated substrates 88 and 89 were investigated by Stoltz and co-workers (Scheme 26). [44] The catalyste mployed possessed the weakly coordinating trifluoroacetate anionsa nd the reactionw as found, in both cases, to follow the syn-pathway, as revealed by the b-hydride elimination of 90 and 91 to give 92 and 93,r espectively.T hese results are in full agreement with the first reaction shown in Scheme 25. Notably,t he catalyst in Scheme 26 allowed the use of molecular oxygen as the stoichiometric oxidant, avoiding the requirement to employ am ediator,s uch as Cu II or quinone/metal macrocycle. [44] The use of the bipyridine ligand presumably keeps the reduced palladium in solution and gives it time to be reoxidized. This study clearly demonstrates that for intramolecular alkoxypalladation,c oordination of the neighboring hydroxy group is apowerful process that favors syn-addition. When the alkoxy group is first coordinated to palladium in the p-olefin complex of 88 and 89,t here will be no competing pathway via anti-attack.
Solvents (e.g.,M eOH vs MeCN) have been found to have an effect on the regiochemistry of the b-hydride elimination but that study did not address the stereochemical issues. [45]

Intramolecular Oxypalladation Followedby Heck Addition
Phenolico lefin 94 has been reported to undergo aP d-catalyzed cyclization (with p-BQ as the terminal oxidant) in the presence of Michael acceptors, such as methyl vinyl ketone, to produce 96 with high enantioselectivity,o wing to the chiral ligand 97 (Scheme 27). Although the steric course of the initial step (94 ! 95)h as not been investigated, it can be assumed to proceed via a syn-mechanism, owing to the weakly coordinating anion (CF 3 CO 2 À )a nd the presence of 97,w hose ligating properties are likely to mirror those of the bipyridine, featured in Scheme 26. [46]

Intramolecular Oxypalladation Followed by Arylation
The Pd II species 100,g enerated in the catalytic cycle from the Pd 0 complex 98 on an oxidative addition to aryl halide 99,h as been shown to react with the alkoxide generated from the bishomoallylic alcohol 101,g iving rise to the cyclization product 104 (Scheme 28). [47] The corresponding alkenyl aryl ether that would be normally expected for these conditions (typical for the Hartwig-Buchwald coupling) has not been observed. The reaction has been suggested to proceed via the h 2 -complex 102 (with pentacoordinated Pd), predisposed to the syn-cyclopalladation, generating the tetrahydrofuran intermediate 103.T he latter speciest hen undergoes re-Scheme26. syn-Mechanismint he intramolecular hydroxypalladation.
Scheme27. Pd II -Catalyzed ring closure followed by Heck addition.
Scheme28. Pd 0 -Catalyzed ring closure followed by arylation. Review ductive elimination to afford 104 as the final arylated product, thereby regenerating Pd 0 98 for the next catalytic cycle. Aldehyde 106 has been isolated as ab yproduct, [47] apparently arising from 105 via a b-hydride elimination.
The stereochemistry of the latter reaction was investigated with the aid of the cis-alkenol 107 (Scheme 29). [48] As expected (based on the discussion in the previous paragraph), the reaction proceeded as a syn-addition to produce 108.H owever,b y using DPPE as as trongly chelating ligand, which apparently prevents the alkoxide coordination to palladium, the mechanism was pushed toward anti-addition,g iving rise to the diastereoisomeric product 109.A mine 110 reactedi nt he same way. [48] The syn-addition mechanism has been masterly utilized in the cyclization,w here the stereochemistry was controlled by the residing chiral center in the secondary alcohol 111 (Scheme 30). [48] Here, the initial oxidative addition to aP d 0 catalyst produces 112,w here the C=Cb ond is coordinated to Pd. Twod iastereofacial ways of this coordination can be considered, so that am ixture might be expected. However,t he subsequent replacement of bromide in the Pd coordination sphere with the neighboring alkoxy group (to generate 113)i s apparently faster for one stereoisomer (as, e.g.,i nS cheme 21), which in view of the reversibility of the h 2 -coordination results in dynamic stereodifferentiation. [49] Hence, isomer 114 is thus formed preferentially and the subsequent reductivee limination affords the cyclization product 115 of high diastereopurity. [48]

Intramolecular Amidopalladation
The first amidopalladation, analogous to alkoxypalladation, was explored in as toichiometric experiment with enamide 116 and found to produce the azepine derivative 117 with the new CÀNa nd CÀPd bonds related in a cis-fashion, which correspond to syn-addition (Scheme 31). [50]

Intramolecular Amidopalladation Followedbyb -Hydride Elimination
Assuming as imilar mechanism for the Pd II -catalyzed cyclization of the alkenyl sulfonamide 118 (Scheme 32), two transition states 120 and 121,d iffering in the mode of Pd coordination can be proposed. The former species (with Pd coordinated to the nitrogen) was found to be much lower in energy than the latter by quantum chemistry calculations. [51,52] Further investigation revealed that the stereochemistry of amidopalladation is subjected to delicate effects of the anion in the originalP dX 2 catalyst andl igand (Scheme 33). [52,53] With the monodeuteriated substrate 122,t he cyclization was catalyzed by two Pd II salts in the presence or absence of ligand 125 and the products were analyzed for the content of deuterium in the resulting cyclic olefins 123 and 124.The latter analysis demonstrated that the chelation of Pd (and thus the synpathway) is suppressed when the combination of (CF 3 CO 2 ) 2 Pd and ligand 125 is used, whereas the remaining reactions were dominated by the syn-pathway. [53b] To add to the complexity of the mosaic, the cyclic, stereospecifically deuteriated sulfonamide 126 was found to be cyclized via the anti-pathway( with 127 and 128 as intermedi-Scheme29. Stereochemistry variation in the Pd 0 -catalyzed ring closure followed by arylation.
Chem. Eur.J.2015, 21,36-56 www.chemeurj.org ates), giving rise to 129,w hich lacks the label (Scheme 34). [54] In this instancet he reaction was catalyzed by Pd(OAc) 2 and would be expected to occur via the syn-addition pathway according to Scheme 33. However,i nt he previouss tudy the final oxidantw as molecular oxygen, whereas in Scheme 34 it was p-BQ, which is known to coordinate to Pd in an h 2 -intermediates [55] and this should be regarded as ac ontributing factor. Also, the fact that the reaction was run under slightly acidic conditions mayi nhibit coordination of the tosylamide. Note that as toichiometrice xperiment with the cyclopentene analogue of 126, N-coordinated to (tBu 2 bipy)PdCl, has been found to cyclize via the syn-pathway in DMSO in the presence of molecular oxygen (to simulate the catalytic conditions). [53a] Clearly, more studies are required to define the relationship between the reaction conditions and the mechanism.
The syn-mechanism has been assumed (but not proven) for the cyclization of 130 [56] and 131 [57] (Scheme 35), where the chiral control is exercised by ac hiral sulfinimide group and chiral ligand 132,r espectively.
Finally,t he O-allyl hemiaminal 137,p repared from the corresponding allylic alcohol and AcOCH 2 NHCbz,h as been shown to undergo an analogous cyclization, giving rise to the transconfigured oxazolidine 138 (Scheme 37) but again without specification as to the actual stereochemistry of the cyclopalladation step. [59] This strategy has been developed as an approach to 1,2-syn-amino alcohols and employed in the synthesis of acosamine. [59]

Intramolecular Amidopalladation FollowedbyA rylation
The cascade of catalytic amidopalladation of olefins and arylation has been developed in parallel with oxypalladation (cf. Section4 .4). Thus, the Pd 0 species, generated from Pd(OAc) 2 and the diphosphine ligand 139 (Scheme 38), [60] has been shown to catalyze the cyclization of the Boc-functionalized aminoalkene 142 to produce the piperidine derivative 145 with high diastereoselectvity.T he latter outcomeh as been rationalized in as imilar way as in the case of oxypalladation, namely by the initial oxidative addition to generate the Pd II complex 141,w hose reaction with 142 (upon deprotonation with Cs 2 CO 3 )g enerates the PdÀNc omplex 143,w here Pd is also coordinated to the C=Cb ond. The key addition across the Scheme33. Anion effect on the stereochemistryoft he sulfonamidopalladation.

Review
C=Cb ond thus occurs with a syn-mechanism to generate the h 1 -complex 144,w hich affords the final product 145 by reductive elimination and Pd 0 that enters the next catalytic cycle. [60,61] The cyclization of the isomeric derivatives of hydroxylamine 146 and 148 (Scheme 39) was expected to proceedv ia the syn-mechanism [62] in analogy to the relatedc yclizations highlighted in Scheme 38 (see also the evidencep resented in Scheme 40). The syn-mechanism was than provedb yi sotopic labeling at the terminus of the double bond. Note, however, that while 146 gave the cis-disubstituted isoxazolidine 147 (typicallyw ith ! 20:1 dr), its positional isomer 148 that reacts by forming the CÀO( rather than CÀN) bond, furnished mainly the trans-product 149 (7:1 dr). Thisc hange of stereochemistry was attributed to the steric interference by the Boc group 1,2related to the phenyli n148;atransition state leadingt o149 is believed to be lower in energy than that producing the cisdiastereoisomer (both using the syn-addition mechanism). [62] Stoichiometric experiments (Scheme 40), [63] employing the potassium salt 150,s tereospecifically deuteriated at the terminus of the doubleb ond, and the ArPdBr complex 151,l end furtherc redence to the syn-mechanism (via 152). [63] The same conclusion has been arriveda tf or an intermolecular, stoichiometric amination of (Z)-CHD=CHD with Ph 2 N-[Pd] [64] and for an intramolecular amidation of ad euterated cyclopentene substrate followedb yb-H elimination. [53a] Another piece of complementary evidencew as obtained from the catalytic cyclization of the deuteriated urea derivative 154.T he relative configuration of the major product 155 was found to be consistent with the syn-mechanism of the addition (Scheme 41). [65] On the other hand, cyclization of the deuteriated acetamide 156 a,c arriedo ut in the absence of astrong base, afforded the arylated pyrrolidine derivative 158 (Scheme 42) as ar esult of an anti-amidopalladation (156 a ! 157 a). The Pd II species 157 a thus generated is oxidized with N-fluorosulfonimide to afford the corresponding Pd IV complex.T he latter intermediate then effects an electrophilica ttack on toluene with retention of configurationt op roduce the C-arylated derivative 158 (thus accomplishing aC À Ha ctivation). The stereochemistry of the anti-aminopalladation has been confirmed by as toichiometric experiment,i n which 156 b was converted into the stable bipy complex 159, characterized by NMR spectroscopy. [66] The dramatic difference in the two mechanisms described in Schemes 41 and 42 apparently originates in the actual reactionc onditions:i nt he former case, the amide-type nitrogen in 154 is deprotonated by as trong base, which increases its propensity to coordinate the palladium catalyst. In the latter case the base is absent and the palladium apparently prefers to first coordinate to the C=C bond;t he resulting species then undergoes at raditional attack from the opposite face as in any classical electrophilic addition.
Numerous other examples highlight the application of this palladium-catalyzed intramolecular amidoarylation of alkenes Scheme38. Intramolecular amidopalladation combined with arylation.

Intramolecular Amidopalladation FollowedbyaS econd Amidation
The amine-typenitrogen, being trivalent,offers another dimension in synthetic strategy that is not available to the divalent oxygen:w hile intramolecular alkoxypalladation produces ac yclic ether that cannot be further elaborated on the oxygen, the analogous amidopalladation features an additional N-substituentt hat can be involvedi nt he subsequent events. Thus, the urea derivative 160 has been shown to undergo double cyclization to give the trans-configured bicyclic product 163 (Scheme 43). [71] The stereochemistry of this cascade was rationalized as follows:t he initial amidopalladation (presumably proceeding with the established syn-mechanism) generates the Pd-chelate 161,i nw hich Pd is replaced by bromide from CuBr 2 with S N 2i nversion and the resulting bromo derivative 162 undergoes cyclization with as econd inversion to produce 163. [71] However,a na lternative pathway,i nvolving anti-amidopallada-tion, followed by oxidative cleavage of the CÀPd bond by S N 2 displacementwith the nitrogen, cannot be excluded.
The related urea derivative 164 (with an ester group in the place of ap henylo ft he previouse xample), also underwent as uccessful cyclization (Scheme 44) but producing the cis-derivative 168 (in contrastt ot he trans-isomer resulting from the previouse xample). The discrepancy was reconciled by assuming just one inversion:h ere, complex 165 (analogous to 161), insteado fu ndergoing the Br À initiated inversion, is believed to coordinate CuBr 2 (167), and Pd itself then serves as al eaving group in the ring closure, [72] so that 168 is formed with as ingle inversion. The difference between the reactivities of 161 and 165 has been attributed to the enolate-typee quilibrium 165 Q 166 that is not available to 161. [71] Nevertheless, an alternative anti-amidopalladation, followed by an oxidative cleavage of the C-Pd by CuBr 2 with inversion, can also be considered.
Ad ifferent type of bis-amidationh as been reported for the stilbene-derived bis-sulfonamide 169 (Scheme 45). [73] Here, the first cyclization presumably proceeds with anti-stereochemistry [74] and the arising Pd II s-complex is believed to be oxidized by PhI(OAc) 2 to generate Pd IV ,w hich enables its replacement with the second sulfonamide group to produce 170. [73] The deuteriated substrate 156 b was used again to elucidate the diamidation process that afforded the pyrrolidine derivative 172 (Scheme 46). The latter outcomec orresponds to the initial anti amidopalladation generating the Pd II species 157 b (as in Scheme 42). Thel atter complex then undergoes oxidation with (PhSO 2 ) 2 NF,g iving rise to the Pd IV species 171,w hich is then converted into the final product 172 on reactionw ith the imide anion via an S N 2i nversion. [66] Scheme42. Intermoleculara midopalladation followed arylation with CÀH activation.

Scheme44. Catalytic intramolecular bis-amidation.
Scheme45. Catalytic intramolecular double-amidation.  [75] has been reported to afford products corresponding to the anti-mechanism, irrespective of the configurationo ft he C=Cb ond in the starting molecule.
The reactionc onditions are mild:a tmospheric pressure of CO, room temperature, and CuCl 2 as the terminal oxidant. Both 5exo and 6-exo cyclizations were attained. In the case of the urea derivatives 175 and 176,t rapping of the acyl-Pd intermediate with the second nitrogen was observed in the absence of methanol to give 180 and 181.I nt he latter instance, the chloro derivative 182 was also obtained, apparently as ap roduct of the S N 2s ubstitution of Pd in the intermediate by chloridei on (as in the case of 21 and 161). [75] With af ree hydroxy group in the molecule as in 184,t he initial amidation has been found to continue by lactonization to the neighboring hydroxyl (Scheme 48). [76] The yield of the resulting lactone 186 was maximized by replacing MeOH (as as olventa nd competitor) with AcOH. While the stereochemistry of the initial amidopalladation has not been investigated in this instance, it was assumed to correspond to the anti-delivery.T he diastereoselectivity is controlled by the hydroxy group, presumably by Pd coordination (185).
Finally,t he Boc-protected hydroxylamine derivative 189 has been shown to undergo the cyclization/carbonylation with an overall syn-addition across the double bond, giving rise to the isoxazoline derivative 191 (Scheme 49). [77] Thes tereochemical outcome [78] apparently originates from the initial coordination of the Boc group to Pd II (190), followedb ysyn-amidopalladation and subsequent carbon monoxide cleavage of the PdÀ Scheme46. Intermoleculara midopalladation followed another amidation.

Scheme48. Amidopalladation followed by lactonization.
Scheme49. Amidopalladation of hydroxylamine derivatives followed by carbonylation. carbon bond with retention of configuration. The striking contrast between the stereochemistry of the carbonylative cyclization of the amides shown in Scheme 47 and the latter case is puzzling, as the reactionc onditions are very similar.O ne explanation is that CuCl 2 was used in Scheme 47, whereas Cu(OAc) 2 was employed in the cyclization of 189 (Scheme 49). The use of CuCl 2 leads to ah igh chloridec oncentration, which makes it more difficult for the amide to coordinate to palladium, and hence the anti-pathways hould be favored. Furthermore, the latter reaction was carriedo ut in the presence MeC(OMe) 3 ,w hich could modify the pH of the mixture and thus improvet he propensity of the ONHCO 2 (tBu) group to coordinate to palladium.

Intermolecular Amidopalladation
The intramolecular amidopalladation has been shown to mostly proceed as a syn-addition, which apparently stems from the considerable entropic advantage of the neighboring group over an external nucleophile.T his factor is absent in intermolecular additions, whichm ay change the stereochemistry, as demonstrated convincingly for the Wacker oxidation (see Chapter 2).

Amidoacetoxylation
While terminal olefins have been shown to readily undergo intermolecular Pd-catalyzeda midoacetoxylation, [79] their internal counterparts resistedanumber of attempts. Finally, cis-olefins, such as 192 (Scheme 50), have been successfully converted into the amidoacetoxylation products on reaction with phthalimide in the presence of PhI(OAc) 2 as the oxidizing agent of palladium. The reaction was originally formulated as proceeding via an initial syn-amidopalladation. However,r ecent re-investigation, [80] prompted by the resultso ft he related diamidation (see the next subchapter), demonstrated that the original structuralassignment of the product [81] was incorrect. This finding led to ar evision of the mechanism,a ccording to which the initial amidopalladation of 192,c atalyzed by Pd II ,p roceeds as ap ure anti process (rather than syn)t og enerate the palladium(II) intermediate 193 that is subsequently oxidized by the hypervalent iodine reagent to produce the Pd IV species 194.T he latter reaction prevents the usual b-elimination and is followed by the S N 2r eplacement of palladium with acetate (i.e.,w ith inversion of configuration [82] )t oa fford the final product 195 that is anti-configured (rather than syn), according to X-ray analysis. [80,83] Noteworthy is also the high regioselectivity of this cascade, owing to the preferential attaching of the palladium moiety to the benzylic position (193). [84] By contrast, the behavior of terminal olefins has been shown to be more complicated:T hus, with the aid of the stereospecifically deuteriated olefin 196 (Scheme 51), the reactionw as found not to be stereoselective, giving rise to a~4:3m ixture of diastereoisomeric amidopalladation products( analogous to 195). The latter outcomeh as been attributed to the poor stereoselectivity of the replacement of Pd with acetate in the final step. On the other hand, under aerobic conditions (i.e.,i nt he absence of the hypervalent iodine reagent), the formation of the (E)-enamide 198 was ascribed to the syn-addition,g enerating 197,f ollowed by the ordinary syn-stereoselective elimination of [PdH].H owever,i ti sp ertinentt on ote that the product 198 was isolated in merely 5% yield (together with 70 %o ft he recovered startingm aterial). [80]

Diamidation
In analogy to acetoxyamidation, the recently developed diamidation (Scheme 52) hasa lso been found to be specific to cisolefins, such as 199.T he formation of the anti-configured of final product 202 has been rationalized by the initial anti-addi-Scheme50. Intermoleculara midoacetoxylation of internal olefins.
Chem. Eur.J.2015, 21, www.chemeurj.org tion of Pd II and phthalimide to generate 200,which is then oxidized with the hypervalent iodine reagent to produce the Pd IV intermediate 201.T he final displacement of Pd with ditosyl imide is believed to proceed with inversion (as in the previous case [80,82] )t op roduce the vicinal diamido derivative 202. [85] 7. Overview of the Stereochemical Outcome in Nucleophilic Addition The syn/anti-dichotomy in the nucleophilic additions was discussed mainly for oxygen and nitrogen nucleophiles. The stereochemical trends are summarized in Ta ble 1, which covers cyclization processes (entries 1-28)a nd intermolecular reactions (entries 29-33).
The intramolecular alkoxypalladation-carbonylation cascade with carbon monoxide and alcohol as the stoichiometric reactants (reaction A)p referentially proceeds via the anti-pathway (entries 1-3), except for the examples, where this mechanism is disfavored by geometrical restrictions in the substrate (entry 4). On the other hand, anti/syn dichotomy hasb een observed for the analogous amidopalladation-carbonylation (reaction H;e ntries 27 and 28). Here, the dramatic differencei n the stereochemistry can be tentatively attributed to the difference in the pH (neutral vs acidic), which is likely to influence the coordination capabilities of the participating amidic group.
Stereochemistry of the intramolecular alkoxypalladation-belimination cascade (reaction B;e ntries 5-7) can be controlled by the anion:t hus, with PdCl 2 ,t hat is, with strongly coordinating chlorides, especially in the presence of additional LiCl, the potentialc oordination of the participating alcoholg roup to Pd is disfavored and the reactionp roceeds as an anti-addition (entry 5). On the other hand, with aw eeklyc oordinating anion (BF 4 À or CF 3 CO 2 À ), Pd II can become coordinated to the participating OH group, which results in the preferential syn-addition (entries 6a nd 7), regardless of the oxidizing reagents or solvent. The analogous intramolecular amidopalladation-b-elimination cascade (reaction E;e ntries13-17) also exhibits the syn/ anti dichotomy,d epending on the actual conditions:t hus, the reactions catalyzed by Pd(OAc) 2 or Pd(O 2 CCF 3 ) 2 follow the synpathway (entries 13, 14, and 16);h owever,t he presence of ligand 125 has been found to drive the reaction catalyzed by Pd(O 2 CCF 3 ) 2 toward the anti-mechanism (entry 15). By contrast, this ligand effect was not observed in the case of Pd(OAc) 2 (entry14), whichi sr ather intriguing. The change of the oxidizing agent (p-BQ vs O 2 )a nd the solvent( THF-DMSO vs toluene) and additives, namely AcOH/AcONa, has been found to also drive the reaction toward anti-mechanism (compare entries 14 and 16 with 17). Calculations predict the syn-mechanism (entry 13), which was also observed for the stoichiometriccyclization using PdCl 2 (reaction D;e ntry 12), where the palladated product was isolated.
In analogy,t he intramolecular amidopalladation-arylation cascade (reaction F;e ntries [18][19][20][21][22], also occurring in the presence of ab ase, invariably proceeds as a syn-addition,r egardless of the solvent or the nature of the ligand employed. On the other hand, the reaction that involves oxidation to Pd IV in thes econdstep(required fort he CÀHa ctivationo fthe "nucleophile")h as been shown to proceed with anti-stereochemistry, even when catalyzed by Pd(O 2 CCF 3 ) 2 (entry 23). Theo utcome in the latter case can be attributed to the absence of the base, which renders the participating N-nucleophile less prone to coordination of the Pd II catalyst.
Intramolecular diamidopalladation (reaction G;e ntries 24 and 25) has been found to prefer the syn-mechanism, apparently due to the same effects as those discussed for entries 18-22. Again, this reaction, involving the Pd II ! Pd IV oxidation and proceeding in the absence of ab ase, favorst he anti-pathway( entry 26).
Intermolecular hydroxypalladation clearly prefers the antimechanism, as shown by the recent studies of the Wacker oxidation (Schemes 14-17). In analogy,i ntermolecular catalytic acetoxypalladation (reaction I;e ntry 29) also follows the antipathway.T he same stereochemistry has been demonstrated for the stoichiometric intermolecular aminopalladation with Me 2 NH (reaction J;entry 30).
Intermolecular amidoacetoxylation (reaction K;e ntry 31) and diamidopalladation (reaction G;e ntry 33), catalyzed by PdCl 2 (which disfavors coordinationo ft he nucleophile to Pd), give the anti-addition products. By contrast, the intermolecular amidopalladation-b-elimination cascade (reaction E;e ntry 32), catalyzed by Pd(OAc) 2 ,f avors the syn-mechanism, which seems to be the only experimentally proven example of syn-migration in intermolecular nucleopalladation with O-and N-nucleophiles (although only in 5% yield) to date. Note that this reaction proceeds with ad ifferent oxidizing agent( O 2 )t han those cited in entries 31 and 33, and that it does not involve the Pd IV species. This, however is unlikely to have am ajor effect on the mechanism of the first step;i ta ppearst hat the key point here is the use of Pd(OAc) 2 rather than PdCl 2 as the catalyst.

Conclusion
This review hasd iscussed the stereochemistry of the palladium-catalyzed addition of nucleophilest oa lkenes anda pplication of these processes in organic synthetic transformations. The syn/anti-dichotomy in the nucleophilic additions was discussed, mainly with oxygen and nitrogen nucleophiles. Ag eneral picture is emerging that in intermolecular reactions the anti-addition of the oxygen and nitrogen nucleophiles to (alkene)Pd II complexes is strongly favored. However,i ni ntramolecular reactions as pecial situation arises when the nitrogen or oxygen nucleophile coordinates to Pd II .I nt his case there is no nucleophile availablef or externala ttack sincet here is a1 :1 ratio between the nucleophilics ite and substrate. Therefore, the syn-attack is tremendously favored in the intramolecular cases where the nucleophile is coordinated to the metal. However,s tereochemistry of the intramolecular reactions is dependent on the coordination capability of the internal nucleophile, which can be modified by the reactionc onditions, so that the whole process can be driven either to the syn-o ranti-pathway. In the intermolecular process, there will alwaysb eaconsiderable amount of free nucleophile in solution and therefore coordination of the nucleophile does not shut down the external anti-pathwaya si sd one in the intramolecular case. As ar esult, external anti-attack is the predominant pathway in the intermolecular nucleophilic addition to (alkene)Pd II complexes.