Evaluation of Pd→B Interactions in Diphosphinoborane Complexes and Impact on Inner‐Sphere Reductive Elimination

Abstract The dative Pd→B interaction in a series of RDPBR’ Pd0 and PdII complexes (RDPBR’=(o‐PR2C6H4)2BR’, diphosphinoborane) was analyzed using XRD, 11B NMR spectroscopy and NBO/NLMO calculations. The borane acceptor discriminates between the oxidation state PdII and Pd0, stabilizing the latter. Reaction of lithium amides with [(RDPBR’)PdII(4‐NO2C6H4)I] chemoselectively yields the C−N coupling product. DFT modelling indicates no significant impact of PdII→B coordination on the inner‐sphere reductive elimination rate.


Introduction
Z-type acceptor ligandsh ave attracted considerable attention over the past decade. [1] Their coordination to transition metals grants access to complexesw ith unusual coordinationg eometries [2] and electronic properties by formation of dative M!Z bonds. Group 13 acceptor ligands, with as pecialf ocus on boranes, have been particularly wells tudied. M!Zb onds can stabilize low oxidation states at the coordinated transition metal. [3] Thus, facile access to complexes featuring transition metals with formally negative oxidationss tates is realized (Figure 1a). [4] Thiss tabilization of low oxidation states appears to inhibit oxidative addition reactions. [3b, e, 5] However,w ed emonstrated that this obstacle can be overcome for complex 1 by addition of catalytic amounts of acetate,w hich competes with Pd 0 for the free coordination site at the borane, thusr eversibly breaking the Pd 0 !Bi nteraction ( Figure 1b). [3b] This concept allowed for the applicationo f1 in catalytic allylic amination, and most recently of 2 in the catalytic hydro-/deutero-dechlorination of aryl chlorides. [3e] Alternatively,b ifunctional substrate activation across the M!Zi nteraction has been described. [3a, 6] The aptitude of hydride, [7] halide [8] and carbon group [9] migration between the Z-type ligand and the coordinatedt ransition metal has initiated further applications. Catalytic processes have concentrated on transformations in which the catalyst is not required to change its oxidation state quickly, butr ather profits from an electronic fine-tuningby electron-withdrawing Z-ligand coordination. [10] Successful applicationsi nclude CO 2 hydrogenation [11] and hydrosilylation, [3d, 12] enyne cycloisomerization [13] and alkyne hydroamination. [14] Michaelis used the heterobimetallic Ti IV /Pd II complex ( Figure 1c), developed by Nagashima, [15] for allylic aminationo fa llyl chlorides with hindered secondary amines. [5b, 16] Combined experimental and computational investigations indicated ar ate enhancement of 10 3À -10 5 of the outer-sphere reductiveC ÀNbond elimination, due to the electron-withdraw- ing Pd II !Ti IV interaction. [5b, 17] This resulta grees with previous investigations performed with Pd h 3 -allyl and Ni h 3 -allyl complexes, which showedf avored reductive outer-sphere reductive elimination in the presence of less electron-donatings pectator ligands. [18] We speculated that the electron-withdrawing propertieso f the borane functionality in diphosphinoborane (DPB) ligands enhances the rate of inner-spherer eductivee limination from Pd complexesd ue to 1) overall reduced electron density at the Pd II center and 2) increasing of the Pd!Bi nteraction strength during reductive elimination. We determineh ow the oxidation state of Pd and co-ligands affect the strength of the Pd!Bi nteraction in DPB complexes. NBO/NLMOc alculations and solidstate structures are used to assess the strength of Pd!Bi nteractions.T he value of the 11 BNMR chemical shift as ap robe is discussed. The reductivee limination of N,N-dimethyl-4-nitroaniline from [( Ph DPB Ph )Pd II (4-NO 2 -C 6 H 4 )NMe 2 ]( 5)w as studied and modelled with DFT calculations to investigate the assumed influence of the borane acceptor.

Syntheses and reactivity of [(DPB)Pd] complexes
As eries of [( Ph DPB Ph )Pd II ]c omplexes was synthesized to examine ap ossible correlation between the nature of ligandsa tP d and the strength of the Pd II !Binteraction (Scheme 1).
Complex [( Ph DPB Ph )Pd II Cl 2 ]( 7)w as produced by reaction of Ph DPB Ph ligand with [(cod)PdCl 2 ]i nD CM and was isolated in 74 %y ield (Scheme 1). Singlec rystalsw ere grown from CH 2 Cl 2 / benzene and analyzed by X-ray diffraction (Figure 2). At ypical square-pyramidal coordination aroundt he palladium was observed around the Pd II center. The chloride ligands are located in cis-configuration at the basal position, and the borane adopts the apical position. The Pd,B distance of 2.762(3) is shorter than the sum of the van der Waals radii (3.28 ), [19] but elongated compared to the sum of the covalent radii (2.23 ). [20] Al ong Pd,C51d istance of 3.405(3) seems to rule out a h 2 -(B,C) type coordination to the Pd II center.As lightly in-creasedp yramidalization at the boron atom is observed (SB a = 355.48)c ompared to complex [( iPr DPB Ph )PdCl 2 ]( SB a = 359.98). [21] The ligand backbone is twisted (dihedral angle C62-C61-C71-C72:3 5.6(3)8)t oa llow for aP -Pd-P angle of 95.49(3)8.T his twist renders the two phosphine groups diastereotopic. The 31 PNMR spectrum of 7 in CD 2 Cl 2 displays two broad resonances of equal integral at d = 39.0 and 48.2 ppm. As eries of 31 P VT NMR spectra was recorded ( Figure 3), covering at emperature range from À29.8 to 35.1 8C. The two singlet resonances coalesced into as ingle resonance (d = 48.2 ppm) at elevated temperatures. The rate constants of the dynamic process were determined by line-shape analysisu sing Bruker's TopSpin software. An Arrheniusp lot analysisg ave an activation energy of E a = 9.3 AE 0.5 kcal mol À1 with ap re-exponential factor of A = (14 AE 7) x 10 9 .
We suggest that the observed dynamic process in the 31 PNMR spectrum of 7 is caused by an interconversion of 7 with its enantiomer ent-7 (Scheme 2).
In order to accommodate for the small P-Pd-P angle of 95.49(3)8,the s-symmetric Ph DPB Ph ligand is twisted. As ar esult, its BÀPh group points towards oneo ft he two phosphine groups,r endering them chemically inequivalent.T his assumption is in line with the observed two 31 PNMR resonances at low temperatures. Twisting of the C62-C61-C71-C72 dihedral angle converts 7 into its enantiomer ent-7,p resumably via a ssymmetric transition in whichthe BÀPh group is orientated between the two chloro ligands.
Complex 8 was synthesized in the same fashion as 7 from [(cod)PdBr 2 ]a nd was isolated in 67 %y ield. The 31 PNMR spectrum displays two broad resonances of equal intensity at d = 45.2 and 38.1 ppm (CD 2 Cl 2 ), suggesting as imilard ynamic process as in 7.D ue to the poor solubility of both 7 and 8,n o 11 BNMR spectra could be obtained. Cationic complex [( Ph DPB Ph )Pd II Cl]SbF 6 (9)w as produced in 51 %i solated yield by halide abstraction from 7 with AgSbF 6 (Scheme 1). Single crystals wereg rown from CH 2 Cl 2 /hexane and analyzed by X-ray diffraction( Figure 2). In the solid state a chloro-bridged dimer [( Ph DPB Ph )Pd II (m-Cl)] 2 (SbF 6 ) 2 is observed with an inversion centerb etween the two Pd II centers. Within the dimer,t he Pd II centeri scoordinated in as quare-pyramidal fashion with the borane located in the apical position. TheP d, Bd istance in complex 9 is 2.721(5) ,w hich is slightly shorter than in [( Ph DPB Ph )Pd II Cl 2 ] 7 (2.762(3) ). However,p yramidalization of the borane is almost identical (SB a = 355.88). Thea bsence of ar elevant h 2 (B,C)!Pd II interaction is suggested by the long Pd1,C1 distance of 3.338(4) .T he Pd,B distance and lack of significant pyramidalization at the borane suggesta weak Pd II !Bi nteraction, which is in line with ab road resonance in the 11 BNMR spectrum at d = 65 ppm (w 1/2 = 1900 AE 500 Hz).
Cationic allyl complex [( Ph DPB Ph )Pd II (h 3 -C 3 H 5 )]SbF 6 (10)w as synthesized by reactiono fA gSbF 6 with zwitterionic allyl com-plex [{(o-PPh 2 C 6 H 4 ) 2 B(OAc)Ph}Pd II (C 3 H 5 )] (4)( Scheme 1) and was isolated in 38 %y ield by crystallization from CH 2 Cl 2 /hexane. Figure 4depictsits solid-state structure. The Pd II centerincomplex 10 is located in at rigonal-pyramidal environment in which the borane occupies the pseudo-apical positiona nd the C 3 H 5 -ligand and the two phosphines are located in the trigonal-planar positions. Aw eak Pd II !Bi nteraction is indicated by aP d,B distance of 2.676(5) ,w hich is in line with am inor pyramidalization at the borane center( SB a = 354.78)a nd ab road 11 BNMR resonance at d = 62 ppm (w 1/2 = 1200 AE 100 Hz). A large Pd,C22d istance of 3.066(6) eliminates the possibilityo f as trong h 2 (B,C)!Pd II interaction. The h 3 -coordinated C 3 H 5ligand is disordered. Using the borane as ar eference point,a 39:61 mixture of the exo-a nd endo-isomers is observed. A wider P-Pd-P angle of 102.86(5)8 is realized by ad ecreasei n the twisting of the ligand backbone (dihedrala ngle C18-C17-C28-C33 of 24.048). The observed disorder of the C 3 H 5 -ligand is in good agreement with the observed NMR spectra.I nt he 31 PNMR spectrum (CD 2 Cl 2 ), two singlet resonances are observedi na40:60 ratio (d = 28.1 and 26.9 ppm) and two sets of C 3 H 5 -units are detected in the 1 HNMR spectrum. DFT calculations (BP86/def-SV(P)) based on the solid-state structures of 10-endo and 10-exo indicateasmall Gibbs free energy preference of DG = 0.74 kcal mol À1 for 10-endo,p redicting a2 9:71 ratio at 298 K.
Complex 5 reacted in as imilarf ashion with LiNCy 2 (26 % 6 after 3h)a nd LiNHtBu (14 % 6 after 5.5 h). However,t he reaction proceeded slower with these sterically more demanding substrates. The reactiono fc omplex 5 with LiNHtBu was monitored for 96 hb y 31 PNMR spectroscopy (46 %c onversion towards 6)w ithout any side products being observed (cf. Ta ble S1). This is in line with the assumption of ar ate-determining transmetalation followed by aq uick reductive elimination.

Analyses of Pd!Binteractions
The solid-state structures of Pd 0/II DPB complexes were analyzed to identify factorsw hich affect the strength of Pd!Bi nteractions. In addition to the new Pd complexes presentedi n this work (6-10) (2) ). Remarkably,e ven the generation of cationic Pd II complexes (9 and 10) has no significant impact on the strength of Pd II !Bi nteractions. The oxidation state at Pd is unambiguously the dominant factor for the strength of the Pd,B bond.
The Pd!Bi nteractions were further analyzed using QM calculations.C omplexes 1, 3, 5-11 and 13 were geometrically optimizedu sing Turbomole 7.0.1 (BP86/def-SV(P)). Ag ood agreement was observed between the optimized structures and their corresponding solid-state structures (Table 1). Complexes 6 and 8 were constructed based on the solid-state Scheme3.Reductiveelimination from 5 and independents ynthesis of 11. structureo fc omplexes 1 and 7.T he Pd!Bi nteractions were furthera nalyzed using NBO/NLMOc alculations. In all cases, an NBO donor/acceptor interaction was foundb etween an occupied d-orbital at Pd and an unoccupied p-orbital at B ( Figure 5). For all examined complexesn orelevant h 2 (B,C)-coordination was found in the NBO calculations.T he Wiberg bond index for Pd,C ipso was below 0.02, with the exception of Pd 0 complexes 1 (0.0697) and 6 (0.0325). Reactivity studies of [(DPB)Pd]-complexes presentedi nt his paper thus appear to be unaffected from significant h 2 (B,C)-coordination. The NBO stabilizing energy of this Pd!Bi nteraction varied depending on the Pd oxidation state. For Pd II !Bi nteractions, an arrow range of NBO stabilizing energiesb etween 8.04 and 11.46 kcal mol À1 waso bserved.S urprisingly,g eneration of cationic complexes (9, 10-endo), exchange of chloro-ligands by bromide( 8)o ri odide/aryl (5)h ad very little effect. In the case of Pd 0 !Bi nteractions,s ignificantlyh igher NBO stabilizinge nergies of 19.53-46.83 kcal mol À1 were found. Regardless of the oxidation state at Pd an approximately linear correlation between the Pd,B distance and the NBO stabilizing energy (E 2 )a ssociated witht he Pd,B interaction was observed ( Figure 6) for 16 valencee lectron (VE) complexes 1,5,7,8,10 and 13.T he Pd,B distance appears to be dictated by the Pd,B bond strength,a nd not by constraints imposed by the chelating ligand.S ubstitution of PPh 2 -groups (6)b yP Cy 2 -groups (3)h ad only am inor effect. The E 2 values for the Pd 0 !Bi nteraction in the 14 VE complexes 3 (46.83 kcal mol À1 )a nd 6 (42.12 kcal mol À1 )s ignificantly deviatef rom this correlationa nd are almostt wicea sm uch as for 16 VE complexes 1 (23.46 kcal mol À1 )a nd 13 (19.53 kcal mol À1 ). Neither the 11 BNMR chemical shift, Pd,B distance or pyramidalization at Bi ndicate ac hange of the Pd 0 !Bi nteraction strength in this magnitudeb etween the 14 VE and the 16 VE complexes (Table 1). This discrepancy mightb ee xplained by the difficulty to comparet he 2 nd order perturbation interaction energies from NBO analysisf rom 14 VE with 16 VE complexes.
The 11 BNMR resonances are shifted linearly towards higher field with an increasing Pd,B distance for Pd 0 complexes,r egardlesso ft he valence electron count at the Pd center ( Figure 6). Complex [( Ph DPB Ph )Pd 0 (PPh 3 )] (2)r eported by Kameo and Bourissou [3e] also fits perfectly into this correlation (d(Pd,B) = 2.294(2) , d( 11 B) 27 ppm). In contrast, the 11 BNMR resonance shifts linearly towards lower field with an increasing Pd,B distance in case of Pd II complexes. 11 BNMR spectroscopy therefore can be used as at ool to assess the strength of Pd! Bi nteractions within ag iven ligand system, providedt hat the oxidation state at the Pd center is taken into account.H owever,g iven the difficulty to determine the precise d( 11 B) of [(DPB)Pd II ]c omplexes (poor solubility and w 1/2 > 1000 Hz ), a certain error for weak Pd II !Bi nteractions needs to be factored in. [26] Quantum chemical calculations (DFT) were used to model the inner-sphere reductivee limination of N,N-dimethyl-4-nitroanilinef rom complex 14-B (Scheme 4). CÀNb ond formation is predicted to proceedv ia an inner sphere reductive elimination with al ow activation barrier of DG°=+7.90 kcal mol À1 (transition state 15-B), yieldingP d 0 complex 6 and N,N-dimethyl-4-nitroaniline (overall DG = À58.75 kcal mol À1 ). In order to understand how the Pd II !Binteraction affects the reductive elimination, the reactionwas also modeled for bis [(2-diphenylphosphino)phenyl]ether (DPEphos) complex 14-O and diphosphinoamine complex 14-N.D PEphos is well established as an effective ligand in palladiumc atalyzed Buchwald-Hartwig-type coupling reactions, [27] and commands very similars tructural features to Ph DPB Ph (Table 2). However,D PEphosc annotm imic  the potentials teric effect of the BÀPh group on the coordinated reactive ligands. For this reason,t he diphosphinoamine ligand (o-PPh 2 C 6 H 4 ) 2 NPh [28] has also been included in the theoretical considerations, as its N-Ph bridgehead gives ag ood model of the B-Ph group in 14-B.E limination of N,N-dimethyl-4-nitroaniline from complexes 14-O and 14-N gave very similar Gibbs free reaction energies of DG = À38.52 kcal mol À1 and DG = À38.63 kcal mol À1 ,r espectively.N oP d 0/II !Ei nteractions were observed in complexes featuring DPEphos and the diphosphinoamine ligand (Table 2, WBI(Pd,E) = 0.005, E = O, N). Given the high structurals imilarity of complexes 6, 16-O and 16-N the increase of DG by ca. 20 kcal mol À1 in case of the Ph DPB Ph ligand is ag ood approximation for the increase of the Pd 0 !Binteraction strengthin6 compared to the Pd II !Binteraction strength in complex 14-B.W hen switching from Ph DBP Ph to DPEphos,asmall decrease of DDG°= 0.41 kcal mol À1 was found for the reductivee limination barrier( Scheme 4). This was surprising, as am ore facile reductivee limination was expected from 14-B than from 14-O,d ue to 1) an electronic effect by Pd!Bc oordination and 2) increased steric bulk of the DPB ligand imposed by the B-Ph group. In case of diphosphinoamine complex 14-N the reductive elimination barrier decreased to DG°= 5.54 kcal mol À1 (DDG°= 2.46 kcal mol À1 ), possibly as ar esult of the increased stericp ressure imposed by the N-Ph group (Table 2) To rule out effects originating fromr estraints imposed by a chelating ligand frame work, the reductive elimination of N,N-dimethyl-4-nitroaniline was also modeled using cis-[(PMe 3 ) 2 Pd II (4-NO 2 C 6 H 4 )NMe 2 ]( 17, DG = 37.47 kcal mol À1 )a nd its Scheme4.Reductiveelimination of N,N-dimethyl-4-nitroaniline from PEP complexes 14-B, 14-O and 14-N.

Conclusions
The strength of Pd!Bi nteractions in [(DPB)Pd] complexes depends primarily on the oxidation state of Pd. In contrast, modificationso ft he DPB ligand or co-ligands have only am inor effect. 11 BNMR spectroscopyh as been established as au seful tool to assess the strengtho fP d !Bi nteractions in solution.
Reactiono fl ithium amides with [( Ph DPB Ph )Pd II (4-NO 2 C 6 H 4 )I] (5) chemoselectively yields the C-N coupling product and [( Ph DPB Ph )Pd 0 ] (6). Inner-sphere reductiveC ÀNb ond elimination was modelled with DFT methods for the Ph DPB Ph ligand. In contrast to reports on acceptor promoted outer-sphere reductive CÀNb ond elimination, [5b, 17] no significant effect of the borane acceptoro nt he inner-sphere reductive elimination rate was found. This is explained by the fact that the strengthening of the Pd!Bb ond occurs after the reductive elimination.

Experimental Section
General All manipulations were performed under an argon atmosphere using standard Schlenk line and glovebox techniques. Glassware was oven dried at 120 8Co vernight and dried with ah eat gun under vacuum prior to use. Te trahydrofuran was dried by an MBraun solvent purification system. Benzene and n-hexane were dried over sodium, distilled under argon prior to use and stored over activated molecular sieves (4 ).
CD 2 Cl 2 and C 6 D 6 were degassed employing the freeze-pump-thaw technique and stored over activated molecular sieves (4

Reactivity studies
As olution of the respective lithium amide (5.7 mmol, 1.1 equiv) in [D 8 ]THF (0.25 mL) was added dropwise over ap eriod of 4min to a stirred solution of nitroarene complex 5 (5.0 mg, 5.2 mmol, 1.0 equiv) in [D 8 ]THF (0.25 mL). The resulting mixture was stirred for another 5min and then transferred into an NMR tube. Reductive elimination was monitored by 31 PNMR spectroscopy.