Aryl-fluoride reductive elimination from Pd(II): Feasibility assessment from theory and experiment

Published on Web 01/17/2007

Aryl-Fluoride Reductive Elimination from Pd(II): Feasibility

Assessment from Theory and Experiment

Dmitry V. Yandulov* and Ngon T. Tran

Contribution from the Department of Chemistry, Stanford UniVersity,

Stanford, California 94305-5080

Received September 30, 2006; E-mail: yandulov@stanford.edu

Abstract: DFT methods were used to elucidate features of coordination environment of Pd(II) that could

enable Ar-F reductive elimination as an elementary C-F bond-forming reaction potentially amenable to

integration into catalytic cycles for synthesis of organofluorine compounds with benign stoichiometric sources

of F^-. Three-coordinate T-shaped geometry of Pd^IIAr(F)L (L ) NHC, PR₃) was shown to offer kinetics and

thermodynamics of Ar-F elimination largely compatible with synthetic applications, whereas coordination

of strong fourth ligands to Pd or association of hydrogen bond donors with F each caused pronounced

stabilization of Pd(II) reactant and increased activation barrier beyond the practical range. Decreasing donor

ability of L promotes elimination kinetics via increasing driving force and para-substituents on Ar exert a

sizable S_NAr-type TS effect. Synthesis and characterization of the novel [Pd(C₆H₄-4-NO₂)ArL(µ-F)]₂(L )

P(o-Tolyl)₃, 17; P(t-Bu)₃, 18) revealed stability of the fluoride-bridged dimer forms of the requisite Pd^IIAr-

(F)L as the key remaining obstacle to Ar-F reductive elimination in practice. Interligand steric repulsion

with P(t-Bu)₃served to destabilize dimer 18 by 20 kcal/mol, estimated with DFT relative to PMe₃analog,

yet was insufficient to enable formation of greater than trace quantities of Ar-F; C-H activation of P(t-Bu)₃

followed by isobutylene elimination was the major degradation pathway of 18 while Ar/F^-scrambling and

Ar-Ar reductive elimination dominated thermal decomposition of 17. However, use of Buchwald’s L )

P(C₆H₄-2-Trip)(t-Bu)₂provided the additional steric pressure on the [PdArL(µ-F)]₂core needed to enable

formation of aryl-fluoride net reductive elimination product in quantifiable yields (10%) in reactions with

both 17 and 18 at 60° over 22 h.

Introduction

fluorine sources such as alkali metal fluorides with phase-

transfer catalysis (PTC), yet accomplish synthetic tasks ordi-

Fundamental advances in understanding and controlling

reactivity of organometallic complexes have paved the way for

the emergence of versatile cross-coupling synthetic methods, a

plethora of which currently enable efficient synthesis of carbon-

carbon and carbon-heteroatom (XdB, Si, P, N, S, O) bonds

by means of late transition metal catalysis.¹Fluorine stands out

as an attractive target for the development of analogous catalytic

C-F bond-forming protocols, whereas traditional-stoichio-

metric-organofluorine chemistry continues to challenge syn-

thetic chemists and demand specialized expertise.^2,3,4The

prospects of developing catalytic methods of C-F bond

synthesis, specifically those that rely on the more benign of the

narily associated with hazards and expenses of anhydrous HF

(AHF) and/or electrophilic fluorinating reagents, are underscored

by the ever-expanding utilization of fluorine functionality in

biologically active compounds^5,2and materials.³Key to realiza-

tion of such catalytic fluorination approaches is the development

of elementary transition metal-mediated C-F bond forming

reactivity, specifically that compatible with alkali MF/PTC as

the in situ source of stoichiometric F^-. Combination of such

reactivity with the various mechanisms of activation and

transformation of organic substrates established broadly across

transition metal series in the context of numerous catalytic

cycles¹potentially could reduce experimental hazards involved

in syntheses of many organofluorine compounds to a routine

level and improve their accessibility.

(1) (a) Transition Metals for Organic Synthesis: Building Blocks and Fine

Chemicals; Beller, M., Bolm, C., Eds.; Wiley-VCH: Weinheim, 2004. (b)

Applied Homogeneous Catalysis with Organometallic Compounds :

A

ComprehensiVe Handbook in Three Volumes; Cornils, B., Herrmann, W.

A., Eds.; Wiley-VCH: Weinheim, 2002. (c) Handbook of Organopalladium

Chemistry for Organic Synthesis; Negishi, E. I., Ed.; Wiley-Interscience:

New York, 2002. (d) Compounds with Transition Metal-Carbon p-Bonds

and Compounds of Groups 10-8 (Ni, Pd, Pt, Co, Rh, Ir, Fe, Ru, Os);

Lautens, M., Ed.; Science of Synthesis: Houben Weyl Methods of

Molecular Transformations, Vol. 1; Thieme: New York, 2000. (e) Top.

Organomet. Chem. 2005, 14.

(4) (a) Hiyama, T. Organofluorine Compounds: Chemistry and Applications;

Springer: New York, 2000. (b) Hiyama, T.; Shimizu, M. Angew. Chem.,

Int. Ed. 2005, 44, 214-231. (c) Ma, J.-A.; Cahard, D. Chem. ReV. 2004,

104, 6119-6146. (d) Top. Curr. Chem. 1997, 193. (e) Wilkinson, J. A.

Chem. ReV. 1992, 92, 505-519.

(5) (a) Jeschke, P. ChemBioChem 2004, 5, 570-589. (b) Pesenti, C.; Fiorenza

Viani, F. ChemBioChem 2004, 5, 590-613. (c) Ojima, I. ChemBioChem

2004, 5, 628-635. (d) Bo¨hm, H.-J.; Banner, D.; Bendels, S.; Kansy, M.;

Kuhn, B.; Mu¨ller, K.; Obst-Sander, U.; Stahl, M. ChemBioChem 2004, 5,

637-643. (e) Biomedical Frontiers of Fluorine Chemistry; Ojima, I.,

McCarth, J. R., Welch, J. T., Eds.; ACS Symposium Series 639; American

Chemical Society: Washington, DC, 1996.

(2) Organo-Fluorine Compounds; Baasner, B., Hagemann, H., Tatlow, J. C.,

Eds.; Methods of Organic Chemistry (Houben-Weyl), Vol. E10; Thieme

Verlag: Stuttgart, 1999.

(3) Organofluorine Chemistry: Principles and Commercial Applications;

Banks, R. E., Smart. B. E., Tatlow, J. C., Eds.; Plenum Press: New York,

1994.

9

1342

J. AM. CHEM. SOC. 2007, 129, 1342-1358

Aryl-Fluoride Reductive Elimination from Pd(II)

A R T I C L E S

Several distinct mechanistic scenarios in which a metal

mediates the formation of a carbon-fluorine bond have been

categorized,³although definitive mechanistic studies are few.⁶

By far the most common elementary pathway is nucleophilic

substitution at carbon by F^-accompanied to a varying degree

by electrophilic assistance from the metal in abstracting the

nucleofuge. This category spans the full range of mechanisms

from limiting A_N+ D_N,⁷through A_ND_N(S_N2) to limiting D_N

+ A_N(S_N1) that broadly characterize halide exchange fluorina-

tion reactivity of binary metal fluorides with aliphatic, vinylic,

aromatic, and carbonyl substrates.²Performing best in such

reactions under strongly acidic conditions,⁸most commonly used

fluorides²of V^V, Cu^II,II/0, Ag^I, Zn^II, Hg^I,II, Tl^I, Pb^II, As^III, Sb^III,V

are suited ideally for catalysis of halide exchange fluorinations

with AHF medium,²rather than with alkali MF/PTC, which

has been claimed.⁹Although lattice energies¹⁰suggest that some

of the above binary fluorides can be formed by halide exchange

of heavier halide salts with alkali MF, treatment of correspond-

ing oxides/hydroxides with HF remains the most practical

preparative route to these historically stoichiometric fluorinating

agents that is also used to prepare alkali MF.²Well-defined η¹-I

iodoalkane organometallic complexes of Re^Iand Ru^IIdisplay

mechanistically related reactivity in yielding corresponding

fluoroalkanes with alkali MF/PTC among other F^-sources.¹¹

Although competing displacement of intact RI with F^-under-

lines the general challenge of maintaining a substrate between

a strong Lewis acid and a nucleophilic F^-, successful develop-

ment of halide exchange fluorination of aliphatic halides with

TlF catalytic in [Ru^IIClP₂L₂]⁺(L ) P or N donor) complexes

of Togni et al.¹²stresses the existence of a useful range of

compatible catalyst Lewis acidity and F^-nucleophilicity.

X₂catalysts,^15a,bSodeoka^15c,dand Cahard^15e,fdeveloped well-

defined Pd^II, Cu^II, and other metal-based catalytic systems, in

all of which C-F bond formation is believed to result from

direct electrophilic attack at chiral metal-bound enolate inter-

mediates by stoichiometric [N-F]⁺electrophilic fluorinating

reagents. Most recently, Sanford et al. reported on directed

oxidative fluorination of unactivated aromatic and aliphatic C-H

bonds with electrophilic [N-F]⁺reagents catalyzed by Pd-

(OAc)₂,^16aexpanding on the extensive range of well-defined

C-O bond-forming oxidative transformations developed with

the same system.^16b-eUltimately reliant on stoichiometric F₂

and/or AHF, none of the above oxidative C-F bond forming

processes are readily amenable to integration into alkali MF-

based catalytic fluorination cycles. However, combination of a

F^-source with chemical,^17,4,a,b,2electrochemical,^2,3or photo-

chemical¹⁸oxidation continues to offer promising opportunities

for the development of less hazardous alternatives to the use of

elemental F₂.

The only known catalytic fluorination process that converts

alkali MF into “advanced” organofluorine compounds, not

accessible from alkali MF directly, is the carbonylative halide

exchange fluorination of aryl halides (fluorocarbonylation, eq

1),¹⁹performed first under relatively mild conditions by Tanaka

et al.²⁰

cat. Pd^II

MF/PTC _{PR , ∆}8 ArC(O)F + MX (1)

Ar-X + CO +

X*F; M)alkali

3

Subsequent reactivity studies by Grushin²¹suggested reduc-

tive elimination of ArC(O)-F from Pd^IIas the mechanism of

C-F bond formation in this system, by analogy to the reductive

elimination of other carboxylic acids derivatives mediating Pd^II

catalysis of the many related carbonylative transformations of

aryl halides.²²

The unique catalytic fluorination system in eq 1 demonstrates

viability of alkali MF as the in situ source of F^-in a Pd^II-

mediated C-F bond-forming process potentially involving

distinct Pd-F intermediates. Additional elementary C-F bond

Oxidative formation of C-F bonds comprises the other major

metal-mediated mechanistic category. A number of high(est)

oxidation state binary metal fluorides,^2,13including heteroge-

neous NiF_xanode phases featured in electrochemical fluorina-

tion,^2,3effect oxidative fluorinations of a wide range of

hydrocarbons and other substrates, in which MF_nreagent is

commonly accepted to^6c,3effect single electron-transfer oxida-

tion(s), subsequently providing, together with AHF medium

where applicable, F^-to the resulting cationic intermediates.

Following the pioneering work of Togni et al. on asymmetric

fluorination of â-keto esters¹⁴with chiral Ti^IV(TADDOLate)-

(15) (a) Hintermann, L.; Togni, A. Angew. Chem., Int. Ed. 2000, 39, 4359-

4362. (b) Piana, S.; Devillers, I.; Togni, A.; Rothlisberger, U. Angew. Chem.,

Int. Ed. 2002, 41, 979-982. (c) Hamashima, Y.; Yagi, K.; Takano, H.;

Tamas, L.; Sodeoka, M. J. Am. Chem. Soc. 2002, 124, 14530-14531. (d)

Hamashima, Y.; Takano, H.; Hotta, D.; Sodeoka, M. Org. Lett. 2003, 5,

3225-3228. (e) Ma, J.-A.; Cahard, D. Tetrahedron: Asymmetry 2004, 15,

1007-1011. (f) Ma, J.-A.; Cahard, D. J. Fluorine Chem. 2004, 125, 1357-

1361.

(16) (a) Hull, K. L.; Anani, W. Q.; Sanford, M. S. J. Am. Chem. Soc. 2006,

128, 7134-7135. (b) Dick, A. R.; Sanford, M. S. Tetrahedron, 2006, 62,

2439-2463. (c) Deprez, N. R.; Kalyani, D.; Krause, A.; Sanford, M. S. J.

Am. Chem. Soc. 2006, 128, 4972-4973. (d) Desai, L. V.; Malik, H. A.;

Sanford, M. S. Org. Lett. 2006, 8, 1141-1144. (e) Dick, A. R.; Kampf, J.

W.; Sanford, M. S. J. Am. Chem. Soc. 2005, 127, 12790-12791.

(17) Subramanian, M. A.; Manzer, L. E. Science 2002, 297, 1665.

(18) (a) Lai, C.; Kim, Y. I.; Wang, C. M.; Mallouk, T. E. J. Org. Chem. 1993,

58, 1393-1399. (b) Habibi, M. H.; Mallouk, T. E. J. Fluorine Chem. 1991,

51, 291-294. (c) Wang, C. M.; Mallouk, T. E. J. Am. Chem. Soc. 1990,

112, 2016-2018.

(6) (a) San Filippo, J., Jr.; Romano, L. J. J. Org. Chem. 1975, 40, 782-787.

(b) Bloodworth, A. J.; Bowyer, K. J. Tetrahedron Lett. 1987, 28, 5347-

5350. (c) Burdon, J.; Parsons, I. W. Tetrahedron 1980, 36, 1423-1433.

(7) Smith, M. B.; March. J. March’s AdVanced Organic Chemistry: Reactions,

Mechanisms and Structure; Wiley-Interscience: New York, 2001.

(8) For a notable exception, see: Poga´ny, S. A.; Zentner, C. M.; Ringeisen,

C. D. Synthesis 1987, 718-719.

(9) Grushin, V. V. U.S. Patent Application 111244,980, Oct. 5, 2005.

(10) CRC Handbook of Chemistry and Physics, Internet Version 2007 (87th

Edition); Lide, D. R., Ed.; Taylor and Francis: Boca Raton, FL, 2007.

(11) (a) Powell, J.; Horvath, M. J. Organometallics 1993, 12, 4067-4072. (b)

Kulawiec, R. J.; Faller, J. W.; Crabtree, R. H. Organometallics 1990, 9,

745-755.

(12) (a) Barthazy, P.; Togni, A.; Mezzetti, A. Organometallics 2001, 20, 3472-

3477. (b) Barthazy, P.; Stoop, R. M.; Wo¨rle, M.; Togni, A.; Mezzetti, A.

Organometallics 2000, 19, 2844-2852. (c) Barthazy, P.; Hintermann, L.;

Stoop, R. M.; Wo¨rle, M.; Mezzetti, A.; Togni, A. HelV. Chim. Acta 1999,

82, 2448-2453. (d) Becker, C.; Kieltsch, I.; Broggini, D.; Mezzetti, A.

Inorg. Chem. 2003, 42, 8417-8429.

(19) (a) Prichard, W. W. U.S. Patent 2696503, 1954. (b) Mador, I. L.; Scheben,

J. A. U.S. Patent 3452090, 1969. (c) Prichard, W. W. U.S. Patent 3632643,

1972.

(20) (a) Sakakura, T.; Chaisupakitsin, M.; Hayashi, T.; Tanaka, M. J. Organomet.

Chem. 1987, 334, 205-211. (b) Okano, T.; Harada, N.; Kiji, J. Bull. Chem.

Soc. Jpn. 1992, 65, 1741-1743. (c) Hatanaka, Y.; Fukushima, S.; Hiyama,

T. Tetrahedron 1992, 48, 2113-2126.

(21) (a) Fraser, S. L.; Antipin, M. Yu.; Khroustalyov, V. N.; Grushin, V. V. J.

Am. Chem. Soc. 1997, 119, 4769-4770. (b) Grushin, V. V. Chem.-Eur.

J. 2002, 8, 1007-1014.

(13) Dukat, W. W.; Holloway, J. H.; Hope, E. G.; Rieland, M. R.; Townson P.

J.; Powell, R. L. J. Chem. Soc., Chem. Commun. 1993, 1429-1430.

(14) (a) Ma, J.-A.; Cahard, D. Chem. ReV. 2004, 104, 6119-6146. (b) Oestreich,

M. Angew. Chem., Int. Ed. 2005, 44, 2324-2327. (c) Mun˜iz, K. Angew.

Chem., Int. Ed. 2001, 40, 1653-1656.

(22) (a) Komiya, S.; Akai, Y.; Tanaka, K.; Yamamoto, T.; Yamamoto, A.

Organometallics 1985, 4, 1130-1136. (b) Grushin, V. V.; Alper, H. J.

Am. Chem. Soc. 1995, 117, 4305-4315. (c) Grushin, V. V.; Alper, H. Chem.

ReV. 1994, 94, 1047-1062. (d) Grushin, V. V.; Alper, H. Top. Organomet.

Chem. 1999, 3, 193-226.

9

J. AM. CHEM. SOC. VOL. 129, NO. 5, 2007 1343

A R T I C L E S

Yandulov and Tran

forming reactions mediated by transition metals may thus be

envisioned that, taking advantage of the existing mechanisms

by which various organic substrates can be transformed in a

catalytic cycle,¹could ultimately enable their functionalization

into diverse organofluorine products in an expedient and

versatile manner that poses few of the traditional hazards of

organofluorine chemistry. As formulated earlier by Grushin,^21b

reductive elimination of alkyl- and aryl-fluorides is one such

elementary reaction that appears especially promising as a nearly

“drop-in” substitute for the product-forming step in the existing

and widely utilized carbon-heteroatom (X ) N, O, S, P) cross-

coupling protocols.¹As of yet, no examples of such elementary

processes have been described in the literature,²³with the most

recent report by Sanford et al.,^16apossibly signifying its

operation from a Pd(IV) state.

Extensive precedent of closely related R-X reductive elimi-

nation reactivity - in particular with X ) OR,²⁴including most

recently X ) OH,^24has well as with X ) Cl, Br and I,²⁵and

possibly X ) F in the catalytic system of eq 1^19,20,21stogether

with vast existing array of substrate activation and modification

options available for both Pd(0) and Pd(II) states¹makes

palladium(II) coordination environment arguably the most

promising platform on which to develop R-F reductive

elimination reactivity. Pursuing such goals in the context of a

novel catalytic approach to fluoroaromatics, Grushin’s system-

atic studies of organometallic fluorides of Pd(II) produced the

first examples of the surprisingly covalent and stable bonding

between the soft Pd(II) center and the hardest base, F^-, in a

series of isolated and fully characterized molecular com-

plexes.^21,26Remarkably, though exhibiting several unique

reactivity modes,^21a,26a-d,fthe novel Pd(II) fluorides of varied

structure and composition were found invariably stable to

reductive elimination of aryl fluorides, evolving instead under

forcingconditionsviaP-FandP-Creductiveelimination.^21b,26e,g,h

Given the difficulty of direct experimental approaches to

produce the desired R-F reductive elimination reactivity with

Pd(II), in the present study we undertook a computational (DFT)

assessment of thermodynamic and kinetic feasibility of this

process taking place from Pd(II) coordination environments

featured in related catalytic transformations. Predicted absolute

activation and equilibrium thermochemistry suggests that pros-

pects of aromatic C-F reductive elimination from Pd(II) are

far from unreasonable when starting with a three-coordinate

PdArF(L) geometry. Maintaining such unsaturated coordination

environment in practice however, may be the key challenge to

address experimentally: we find {Pd(C₆H₄-4-NO₂)F(P(t-Bu)₃)},

whose Br and I analogs exhibit unique three-coordinate geom-

etries in solution and solid state,^25b,eto exist as a fluoride-bridged

dimer as does the novel [Pd(C₆H₄-4-NO₂)F(P(o-Tolyl)₃)]₂at

temperatures of up to 95 °C in solution. Stability of dimeric

ground states even in these Pd(II) coordination environments

of high degree of steric congestion ultimately limits the extent

of Ar-F reductive elimination to trace levels, with overwhelm-

ing preference displayed for a plethora of conventional C-C,

P-C, C-H, and P-F bond-forming/breaking decomposition

pathways. Nevertheless, both isolated [Pd(C₆H₄-4-NO₂)L(µ-F)]₂

produce, in the presence of Buchwald’s P(t-Bu)₂(C₆H₄-2-Trip),

the sought F-C₆H₄-4-NO₂in 10% yield over 22 h at 60 °C.

Results

A. Computational Studies. 1. Benchmark of the Compu-

tational Models. The computational models used in the studies

of the as-yet unobserved processes of C-F reductive elimination

from Pd(II) environments were tested on (o-Tolyl)-bromide

reductive elimination equilibrium of Pd(o-Tolyl)Br(P(t-Bu)₃)/

P(t-Bu)₃, thermodynamics and kinetics of which have been

determined by Roy and Hartwig^25c,d(Table 1).

Although most geometrical parameters are reproduced fairly

well with both MO and ONIOM models (within 2.6% for

overestimated Pd-P(t-Bu)₃distances), the key difference from

the experiment lies in the substantially longer single Pd‚‚‚HC

agostic bond featured in the optimized geometries. The inad-

equate description of the agostic interaction, which serves to

stabilize Pd(II) state, may be the reason why equilibrium and

activation free energies are underestimated by 0 (6) and 5 (6)

kcal/mol with the best MO (ONIOM) models, which include

solvation energies. It is notable that the Pd‚‚‚HC distance is

shorter in the ONIOM model, which describes the agostic

interaction only approximately with electrostatic nonbonded term

of the UFF energy of the real system, but represents adequately

inter- and intraligand steric repulsion important in pushing the

agostic C-H bond to Pd.²⁸Within the limits of the above

discrepancies, ONIOM performs adequately as compared to MO

models, which reproduce experimental BDE’s of Ph-F (125.6)

and Ph-Cl (95.5)¹⁰as 123.4 and 93.3 kcal/mol, respectively

(B3PW91/BS II //BS I). Being largely equivalent to the BS I

energy/BS I geometry scheme otherwise, use of BS II single-

point energies on BS I-optimized geometries reduced BSSE’s

in the energies of fluoride-bridged dimers to a negligible level

and was therefore chosen as the standard approach. Computa-

tional data presented in the following can thus be considered

accurate to within 5 kcal/mol relative to the experiment, with

(23) ArTl^IIIF₂and ArPb^IV(OAc)₃are known to yield corresponding ArF on

treatment with BF₃, possibly in a manner of concerted reductive elimination

from a high-valent organometallic fluoride intermediate; the incipient aryl

cation character noted of the ArTl^IIIF₂reactivity suggests fluoro-dediazo-

+

niation (ArN₂+ F^-f ArF) as a closer mechanistic analogy for these

reactions than a concerted Ar-F reductive elimination: (a) De Meio, G.;

Morgan, J.; Pinhey, J. T. Tetrahedron 1993, 49, 8129-8138. (b) Taylor,

E. C.; Bigham, E. C.; Johnson, D. K. J. Org. Chem. 1977, 42, 362-363.

(24) (a) Mann, G.; Hartwig, J. F. J. Am. Chem. Soc. 1996, 118, 13109-13110.

(b) Widenhoefer, R. A.; Zhong, H. A.; Buchwald, S. L. J. Am. Chem. Soc.

1997, 119, 6787-6795. (c) Widenhoefer, R. A.; Buchwald, S. L. J. Am.

Chem. Soc. 1998, 120, 6504-6511. (d) Mann, G.; Incarvito, C.; Rheingold,

A. L.; Hartwig, J. F. J. Am. Chem. Soc. 1999, 121, 3224-3225. (e) Aranyos,

A.; Old, D. W.; Kiyomori, A.; Wolfe, J. P.; Sadighi, J. P.; Buchwald, S. L.

J. Am. Chem. Soc. 1999, 121, 4369-4378. (f) Mann, G.; Shelby, Q.; Roy,

A. H.; Hartwig, J. F. Organometallics 2003, 22, 2775-2789. (g) Voro-

gushin, A. V.; Huang, X.; Buchwald, S. L. J. Am. Chem. Soc. 2005, 127,

8146-8149. (h) Anderson, K. W.; Ikawa, T.; Tundel, R. E.; Buchwald, S.

L. J. Am. Chem. Soc. 2006, 128, 10694-10695.

(25) (a) Roy, A. H.; Hartwig, J. F. J. Am. Chem. Soc. 2001, 123, 1232-1233.

(b) Stambuli, J. P.; Bu¨hl, M.; Hartwig, J. F. J. Am. Chem. Soc. 2002, 124,

9346-9347. (c) Roy, A. H.; Hartwig, J. F. J. Am. Chem. Soc. 2003, 125,

13944-13945. (d) Roy, A. H.; Hartwig, J. F. Organometallics 2004, 23,

1533-1541. (e) Stambuli, J. P.; Incarvito, C. D.; Bu¨hl, M.; Hartwig, J. F.

J. Am. Chem. Soc. 2004, 126, 1184-1194.

(26) (a) Pilon, M. C.; Grushin, V. V. Organometallics 1998, 17, 1774-1781.

(b) Grushin, V. V. Angew. Chem., Int. Ed. 1998, 37, 994-996. (c)

Flemming, J. P.; Pilon, M. C.; Borbulevitch, O.; Ya Antipin M. Yu.;

Grushin, V. V. Inorg. Chim. Acta, 1998, 280, 87-98. (d) Marshall, W. J.;

Thorn, D. L.; Grushin, V. V. Organometallics 1998, 17, 5427-5430. (e)

Grushin, V. V. Organometallics 2000, 19, 1888-1900. (f) Roe, D. C.;

Marshall, W. J.; Davidson, F.; Soper, P. D.; Grushin, V. V. Organometallics

2000, 19, 4575-4582. (g) Grushin, V. V.; Marshall, W. J. Angew. Chem.,

Int. Ed. 2002, 41, 4476-4479. (h) Marshall, W. J.; Grushin, V. V.

Organometallics 2003, 22, 555-562.

(28) Ujaque, G.; Cooper, A. C.; Maseras, F.; Eisenstein, O.; Caulton, K. G. J.

Am. Chem. Soc. 1998, 120, 361-365.

(27) Tanaka, M. Acta Cryst. 1992, C48, 739-740.

9

1344 J. AM. CHEM. SOC. VOL. 129, NO. 5, 2007

Aryl-Fluoride Reductive Elimination from Pd(II)

A R T I C L E S

Table 1. Comparison of Various Computational Models in Describing Experimental Structures (Å, deg) and Energetics (kcal/mol) Pertaining

to (o-Tolyl)-Br Reductive Elimination from Pd(o-Tolyl)Br(P(t-Bu)₃)

Pd−P(t-Bu)₃

Pd

−Ar

Pd−Br

CH‚‚‚Pd^a

P

−Pd

−Ar

Ar−Pd−X

∆

G^q

∆G°

+70

+65

°

C

°C

Pd(2,4-Xyl)I(P(t-Bu)₃)^b

PdPhBr(P(t-Bu)₃)^c

Pd(P(t-Bu)₃)₂

2.294(1)

2.2854(13)

2.285(3)

1.982(4)

1.977(5)

2.26

2.13

100.9(1)

99.92(15)

94.0(1)

93.14(5)

2.4537(8)

2.410

d

Pd(o-Tolyl)Br(P(t-Bu)₃)

24.1^e

15.8

15.7

17.8

17.3

17.0

19.2

-0.8 ^f

-8.4

-9.9

-10.2

-7.0

-2.4

-3.8

-1.0

ONIOM^g/BS I//BS I^h

2.344/2.294ⁱ

1.968

1.973

2.41

2.48

103.1

101.7

96.0

94.4

ONIOM/BS II//BS I^j

ONIOM/BS III//BS I^k

l

ONIOM/BS II//BS I + B3PW91/BS I/C₆H₆

B3PW91/BS I//BS I

2.349/2.332ⁱ

2.425

B3PW91/BS II //BS I

B3PW91/BS II //BS I + B3PW91/BS I/C₆H₆

l

^aC-H bonds extended to 1.083 Å in X-ray structures. ^bCCDC structure code PABLOU, ref 25e. ^cCCDC structure code EROGUN, ref 25e. ^dCCDC

structure code TARDUL, ref 27. ^eComputed from the value of 2.0 × 10^-3s^-1for rate constant k₁of path C in ref 25c according to the Eyring theory.

^fComputed from equilibrium constant value of 32.7 × 10^-1of ref 25c. ^gB3PW91/UFF with P(t-Bu)₃represented as PMe₃at the QM level. ^hSDD

quasirelativistic ECP and basis set augmented with polarization functions on P, Br, and Pd, 6-31G(d,p) on C and H. ⁱDistances in Pd(o-Tolyl)Br(P(t-Bu)₃)/

j

Pd(P(t-Bu)₃)₂. Single point on BS I-optimized geometry with BS II: Martin’s extended SDD basis set on Pd and Br, 6-311+G(2d,p) on C, H, P. ^kSingle

point on BS I-optimized geometry with BS III: Martin’s extended SDD basis set on Pd and Br, AUG-cc-pVTZ on C, H, P. ^lCorrected for solvation with

B3PW91/BS I single points on BS I optimized geometry with CPCM, a conductor-like screening model of C₆H₆solvent. See experimental section for

additional details of the computational methods.

deviations being potentially as significant for systems with

agostic interactions,²⁹and solvation contributions being

important.

reaction), together with ability of NHC’s to sustain otherwise

harsh, oxidizing conditions of catalytic methane functionalization

as supporting ligands for Pd.^34cFollowing this reasoning and

taking into account the greater intrinsic propensity of three- vs

four-coordinate d⁰metal centers to undergo related C-X

reductive eliminations,^35,25Pd(Ph)F(Me₂NHC) became the initial

focus of the computational study.

Reductive elimination of Ph-F from T-shaped Pd(Ph)F(Me₂-

NHC) (1) is computed to require ∆H^q) 28.9 kcal/mol in the

gas phase (1-2^q, Figure 1). Similar to the value in solution

(C₆H₆, ∆H^q) 31.2) as well as the ∆G^q) 29.8, the enthalpy of

1-2^qimplies a reaction time scale of hours at 100 °C. Partial

IRC calculations (Supporting Information) confirmed Ph-F

reductive elimination origin of 1-2^q, although could not be

extended to connect the reaction path from 1-2^qdirectly to

either the metastable PhF-κF Pd⁰adduct 2 or another TS (2-

3^q) with imaginary frequency normal mode depicting closely

κFT1,2η isomerization of PhF adducts 2 and 3. These two

stationary points, possibly traversed along the way from 1-2^q

2. Ar-F Reductive Elimination from {Pd(Ar)F(NHC)}.

Pd(Ph)F(Me₂NHC). Of the two common types of auxiliary

ligands broadly able to support both Pd(II) and Pd(0) states in

homogeneous catalysisstertiary phosphines (PR₃) and N-

heterocyclic carbenes (NHC)¹sthe former are susceptible to

F^--mediated intramolecular oxidation by Pd(II), viz. P(III)/Pd-

(II)fP(V)/Pd(0), which yields fluorophosphines (R₂PF), fluoro-

phosphoranes (R₂PF₃, R₃PF₂, R₄PF), and reductively coupled

R-R, R₂P-PR₂products together with reduced Pd.^{30,21b,26e,g,h}

Related to the other hard base-mediated oxidations of PR₃by

Pd(II),³¹the fluoride-induced redox reactivity renders even such

deceptively ordinary complexes as PdF₂(PR₃)₂stable only at

or below 0 °C in solution.³²Although tertiary phosphines

successfully enable late transition metals catalysis of a wide

variety of reactions that involve stoichiometric fluoride,³³strong

preference for P-F over C-F bond formation established for

all known organometallic Pd(II) fluorides by Grushin suggests

that auxiliary ligands other that PR₃may need to be explored.

While only a few M(NHC)F complexes have been characterized

so far,^34aNHC’s could offer improved stability to intramolecular

oxidation by Pd(II)/F^-in view of the potent fluorodehydroxy-

lating reactivity of 2,2-difluoro-1,3-dimethylimidazolidine^34b(a

formal analog of a difluorophosphorane in such oxidation

(33) For example, acting as a desilylating agent in Hiyama coupling^1cand

numerous other circumstances: Schwesinger, R.; Link, R.; Wenzl, P.;

Kossek, S. Chem.-Eur. J. 2006, 12, 438-445.; as a base¹; as a nucleofuge

in cross-couplings with aryl fluorides: Kim, Y. M.; Yu, S. J. Am. Chem.

Soc. 2003, 125, 1696-1697. Saeki, T.; Takashima, Y.; Tamao, K. Synlett

2005, 1771-1774. Yoshikai, N.; Mashima, H.; Nakamura, E. J. Am. Chem.

Soc. 2005, 127, 17978-17979. Bahmanyar, S.; Borer, B. C.; Kim, Y. M.;

Kurtz, D. M.; Yu, S. Org. Lett. 2005, 7, 1011-1014. Terao, J.; Watabe,

H.; Kambe, N. J. Am. Chem. Soc. 2005, 127, 3656-3657. Jasim, N. A.;

Perutz, R. N.; Whitwood, A. C.; Braun, T.; Izundu, J.; Neumann, B.;

Rothfeld, S.; Stammler, H.-G. Organometallics 2004, 23, 6140-6149.

Braun, T.; Perutz, R. N. Chem. Commun. 2002, 2749-2757.; or even as

(29) These are limited to the structures optimized in the benchmark analysis,

_PtBu10_NOand _PtBu10-11_NO^q(vide infra, Figure 9); in all other computed

nucleophile in the original Tanaka’s system (eq 1)²⁰

.

3

2

three-coo²rdinate Pd(II) derivatives and transition states for reductive

elimination of Ar-X auxiliary ligand hydrogens do not approach the metal

closer than 2.7 Å. Cartesian coordinates and drawings with close contacts

highlighted are provided in Supporting Information.

(34) (a) Chatwin, S. L.; Davidson, M. G.; Doherty, C.; Donald, S. M.; Jazzar,

R. F. R.; Macgregor, S. A.; McIntyre, G. J.; Mahon, M. F.; Whittlesey, M.

K. Organometallics 2006, 25, 99-110. Schaub, T.; Radius, U. Chem. Eur.

J. 2005, 11, 5024-5030. Laitar, D. S.; Mu¨ller, P.; Gray, T. G.; Sadighi, J.

P. Organometallics 2005, 24, 4503-4505. (b) Hayashi, H.; Sonoda, H.;

Fukumura, K.; Nagata, T. Chem. Commun. 2002, 1618-1619. (c) Mue-

hlhofer, M.; Strassner, T.; Herrmann, W. A. Angew. Chem., Int. Ed. 2002,

41, 1745-1747.

(30) (a) Mason, M. R.; Verkade, J. G. Organometallics 1990, 9, 864-865. (b)

Mason, M. R.; Verkade, J. G. Organometallics 1992, 11, 2212-2220. (c)

Marshall, W. J.; Grushin, V. V. J. Am. Chem. Soc. 2004, 126, 3068-3069.

(d) Macgregor, S. A.; Roe, D. C.; Marshall, W. J.; Bloch, K. M.;

Bakhmutov, V. I.; Grushin, V. V. J. Am. Chem. Soc. 2005, 127, 15304-

15321.

(31) Grushin, V. V. Chem. ReV. 2004, 104, 1629-1662.

(32) (a) Yahav, A.; Goldberg, I.; Vigalok, A. J. Am. Chem. Soc. 2003, 125,

13634-13635. (b) Yahav, A.; Goldberg, I.; Vigalok, A. Inorg. Chem. 2005,

44, 1547-1553.

(35) (a) Tatsumi, K.; Hoffmann, R.; Yamamoto, A.; Stille, J. K. Bull. Chem.

Soc. Jpn. 1981, 1857-1867. (b) Morvskiy, A.; Stille, J. K. J. Am. Chem.

Soc. 1981, 103, 4182-4186.(c) Driver, M. S.; Hartwig, J. F. J. Am. Chem.

Soc. 1997, 119, 8232-8245.(d) Hartwig, J. F. Acc. Chem. Res. 1998, 31,

852-860.(e) Yamashita, M.; Hartwig, J. F. J. Am. Chem. Soc. 2004, 126,

5344-5345. (f) Hartwig, J. F. Synlett 2006, 1283-1294.

9

J. AM. CHEM. SOC. VOL. 129, NO. 5, 2007 1345

A R T I C L E S

Yandulov and Tran

Figure 1. Computed reactivity profile of Pd(Ph)F(Me₂NHC) and related analogs: Pd(Ph)Cl(Me₂NHC) (numbers in green, right of center or below), Pd-

(Ph)F(4,5-Cl₂-Me₂NHC) (numbers in blue, left of center), Pd(C₆H₄-4-NO₂)F(Me₂NHC) (numbers in red, below center), and Pd(C₆H₄-4-OMe)F(Me₂NHC)

(numbers in cyan, second below center).

to the global minimum Pd⁰product 1,2η-PhF adduct 3, lie well

below the reductive elimination TS 1-2^q(Figure 1) and would

not be kinetically significant in the experimental practice. The

final product of reductive elimination, 3, is nearly isoenergetic

with 1 (∆H° ) +1.2 kcal/mol) and lies 7.2 kcal/mol in ∆G°

below the separated Pd⁰(Me₂NHC) + PhF fragments. Thus, both

kinetics and thermodynamics of Ph-F reductive elimination

from a three-coordinate Pd(II) environment representative of

those mediating closely related C-X bond forming reductive

elimination processes^25,35are computed to fall in a practical

experimental range. Analogous elimination of Ph-Cl from 1_Cl

is characterized by ∆H^q) 26.8 kcal/mol and ∆H° ) +16.5

kcal/mol.³⁶The substantial computed increase in the driving

force for reductive elimination on going from Ph-Cl to Ph-F

is largely attributable to Pd^II-X BDE (∆H°_{Pd-XfPd•+X•}, 1_Cl:

89.5; 1: 103.3 kcal/mol) lagging behind that of Ph-X (Cl: 93.3,

F: 123.4 kcal/mol) on going to lighter halides. Given the

demonstrated precedent of reductive elimination of heavy aryl

halides from Pd(II),^25a,c,dthe magnitude of the computed driving

force increase on going to F offers a promising outlook for

thermodynamics of such Ar-F bond forming reactions.

reactions of M(R)(NHC) intermediates.³⁷Three isomeric forms

of the Ph-NHC migration product were located in the case of

4 within 3.8 kcal/mol of each other, with one (4A) clearly

maintaining Pd(II) oxidation state in a distorted Y-shaped

geometry and the others (4 and 4B) best described by mixed

Pd⁰TPd^IIzwitter-ionic resonance forms also shown in Figure

1.³⁸Although the reaction of 1 rate-limited by 1-4^qis

endothermic by 6.1 kcal/mol (4) and thus may exhibit at least

some degree of reversibility, irreversible decomposition by loss

of Ph-NHCMe₂⁺from 4 and aggregation to Pd black is likely

to be facile enough to preclude 1 from undergoing Ph-F

reductive elimination altogether.

Most importantly, the very existence of three-coordinate 1

in solution requires ∆G° ) 13.5 kcal/mol (per monomer) relative

to the fluoride-bridged dimer form [1]₂(Figure 1). Inclusion of

this uphill pre-equilibrium into consideration places ∆G^qfor

Ph-F reductive elimination chiefly out of practical experimental

range, at 13.5 + 29.8 ) 43.3 kcal/mol for a rapid pre-

equilibrium dissociation of [1]₂followed by rate-limiting

elimination from 1 via TS 1-2^q(complete dimer cleavage could

(37) (a) McGuinness, D. S.; Green, M. J.; Cavell, K. J.; Skelton, B. W.; White,

A. H. J. Organomet. Chem. 1998, 565, 165-178. (b) McGuinness, D. S.;

Cavell, K. J.; Skelton, B. W.; White, A. H. Organometallics 1999, 18,

1596-1605. (c) McGuinness, D. S.; Saendig, N.; Yates, B. F.; Cavell, K.

J. J. Am. Chem. Soc. 2001, 123, 4029-4040. (d) Nielsen, D. J.; Magill, A.

M.; Yates, B. F.; Cavell, K. J.; Skelton, B. W.; White, A. H. Chem.

Commun. 2002, 2500-2501. (e) Caddick, S.; Cloke, F. G. N.; Hitchcock,

P. B.; Leonard, J.; Lewis, A. K. de K.; McKerrecher, D.; Titcomb, L. R.

Organometallics 2002, 21, 4318-4319. (f) Lewis, A. K. de K.; Caddick,

S.; Cloke, F. G. N.; Billingham, N. C.; Hitchcock, P. B.; Leonard J. J. Am.

Chem. Soc. 2003, 125, 10066-10073. (g) Marshall, W. J.; Grushin, V. V.

Organometallics 2003, 22, 1591-1593. (h) Crudden, C. M.; Allen, D. P.

Coord. Chem. ReV. 2004, 248, 2247-2273. (i) Cavell, K. J.; McGuinness,

D. S. Coord. Chem. ReV. 2004, 248, 671-681. (j) Graham, D. C.; Cavell,

K. J.; Yates, B. F. Dalton Trans. 2005, 1093-1100. (k) Graham, D. C.;

Cavell, K. J.; Yates, B. F. Dalton Trans. 2006, 1768-1775.

However, 1 would be considerably more prone to elimination

of phenyl-imidazolium ion, transition state for which (1-4^q) is

computed to lie 8.8 kcal/mol below that for Ph-F reductive

elimination (1-2^q). Extensively studied experimentally and

computationally by Cavell and Yates, such intramolecular

migratory processes resulting in transfer of hydrocarbyl ligands

to NHC carbon are now well-known to limit the usefulness of

NHC’s as supporting ligands to relatively facile productive

(36) ∆G^q(C₆H₆) ) 10.6 kcal/mol for the reverse process, oxidative addition of

PhCl in 3_Cl, compares well with the value of 9.4 kcal/mol obtained for this

process in a different computational study: Green, J. C.; Herbert, B. J.;

Lonsdale, R. J. Organomet. Chem. 2005, 690, 6054-6067.

(38) Contributing resonance forms have been identified on the basis of computed

geometrical parameters only, available in Supporting Information. Mixed

resonance form assignments do not bear any quantitative connotation.

9

1346 J. AM. CHEM. SOC. VOL. 129, NO. 5, 2007

Aryl-Fluoride Reductive Elimination from Pd(II)

A R T I C L E S

but smaller influence on the driving force (∆∆H° ) -2.1 kcal/

mol). Like the related polar carbon-heteroatom reductive

eliminations,^35d,fand to a greater extent nucleophilic aromatic

substitutions,⁷elimination of Ar-F from Pd^IIis promoted by

electron-withdrawing para-substituents of the aryl group, pri-

marily in kinetics. On the contrary, reductive elimination of Ar-

NHC⁺from 1_OMe-1_NOis inhibited by electron-withdrawing

2

p-Ar substituents (∆∆H^q) +4.5 kcal/mol), signifying impor-

tance of a dative interaction between the filled Ar σ orbital and

NHC π* to the bonding in the TS 1-4^q. To an extent, Ar-

NHC⁺elimination from 1 is initiated by a nucleophilic attack

of the Ar at the NHC C-2.^37kThermodynamics of Ar-NHC⁺

elimination (4) vary in the opposite direction from kinetics (1-

4^q) in the same series 1_OMe-1_NO(∆∆H° ) -1.8 kcal/mol)

2

apparently reflecting the dominant contribution of the Pd⁰

resonance form, whose stability is indeed enhanced by stronger

π-acidity of the π-bound 1,2η-C₆H₄-4-X aryl ring, to the

structure of 4. Because characteristics of the aryl groups that

promote Ar-F (1-2^q, electrophilic) and Ar-NHC⁺(1-4^q,

nucleophilic) reductive eliminations from 1 are opposite of each

other, variation of para-substituent from OMe to NO₂causes a

pronounced reduction in the relative preference for (X-4-C₆H₄)-

NHC⁺elimination over that of (X-4-C₆H₄)-F, from 12.4 to

2.9 kcal/mol in ∆∆H^q. Thus, Ar-F elimination may be able to

out compete decomposition of PdAr(F)(NHC) via Ar-NHC⁺

elimination with electron-deficient Ar.

Figure 2. Computed effects of coordination of an additional strong donor

ligand on Ph-F reductive elimination from a three-coordinate PdPh(F)L.

require even greater activation, vide infra). At the same time,

+

[1]₂may still suffer reductive elimination of Ph-NHCMe₂

,

either via 1 and TS 1-4^qor directly from saturated [1]₂, given

such precedent for at least the alkyl-NHC eliminations from

neutral four-coordinate d⁸metal complexes.^37cThus, it appears

that any potential benefit of improved stability to intramolecular

oxidation by Pd(II)/F^-offered by NHC’s in comparison to

phosphines would be undermined by susceptibility of the former

to reductive elimination with the hydrocarbyl group, at least as

long as the ∆H^qfor Ph-F reductive elimination remains

noncompetitive.

Replacement of Me₂NHC with electron-deficient 4,5-dichloro

Me₂NHC in _Cl21 promotes kinetics of Ph-F reductive elimina-

tion (∆∆H ) -0.8 kcal/mol) in accord with Hammond postulate

(∆∆H° ) -1.6 kcal/mol) and exerts nearly opposing influence

on kinetics (_Cl21-4^q, ∆∆H^q0.0) and thermodynamics (_Cl24,

∆∆H° ) +2.1) of Ph-NHC⁺reductive elimination, as expected

for the latter process starting out as a nucleophilic attack of Ar

Pd(Ph)F((MeNHC)₂CH₂). A productive strategy to inhibit

hydrocarbyl-NHC reductive elimination from d⁸metal com-

plexes is to incorporate the NHC ligand in a multidentate

assembly that resists rotation of the NHC ring out of the

coordination plane, into an orientation adequately suited for

formation of partial bonding between the eliminating fragments

(cf. 1-4^q).^37d,iSeeking to assess the utility of this approach with

the chelating bis-carbene ligand (MeNHC)₂CH₂in the compu-

tational model 5 (Figure 2) showed instead a remarkably

profound effect of coordinative saturation on the kinetics of

Ph-F reductive elimination from Pd(II).

at C-2 (_Cl21-4^q, cf. 1-4_OMe^q, 1-4_NO^q, vide supra), but resulting

2

in a net increase of positive charge on the NHC ring in the

product (_Cl24).

The Role of F^-. Increasing electrophilicity of either Pd (1-

2^qvs _Cl21-2^q) or the aryl (1-2_OMe^qvs 1-2_NO^q) causes relative

2

stabilization of 1-2^qas the Ar-F bond-making transition state.

Could the difficulty of making the C-F bond from 1 via 1-2^q

reflect localization of excessive negative charge on F^-in the

TS⁴⁰and may the process be facilitated by binding an external

electrophile directly to F^-? To assess these hypotheses, an HF

molecule was added to the Pd(Ph)F(Me₂NHC) model 1 as a

hydrogen-bond donor to the F^-ligand. Although it was possible

to fully optimize transition state 1-2_...HF^qwith HF bound almost

solely via the (Pd)F‚‚‚HF hydrogen bond, energies of the purely

(F)H‚‚‚F(Pd) hydrogen-bonded HF adducts 1_...HFand 3_...HFwere

obtained as upper estimates from incomplete optimizations due

to their respective instabilities to 1_-HFand 3_-HF(Figure 3). The

comparative electronic energy profile that emerged resembles

that of the bi-/monodentate (MeNHC)₂CH₂models 5, 5_η1(Figure

2): like coordination of the additional ligand to Pd, formation

of a hydrogen bond to F^-undergoing Ph-F reductive elimina-

tion serves mostly to stabilize Pd^IIreactant 1, reducing the

driving force (∆∆E° ) 12.1 kcal/mol) and raising the activation

Coordination of the second arm of the chelating bis-carbene

ligand to Pd, while favorable for all stationary points on the

elimination pathway, stabilizes the oxidized Pd^IIstate (5_η1) by

nearly 22 kcal/mol more than it does both the elimination

q

transition state 5-6_η1and the Pd⁰product 6_η1. The relative

stabilization of the Pd^IIreactant raises the magnitude of the

Ph-F reductive elimination ∆H^qto 50.6 kcal/mol (5-6^q), or

well outside of the realm of synthetically relevant values.

Considered in reverse, increase of the elimination driving force

by destabilization of the saturated reactant 5 via ligand loss is

invaluable in translating into a substantial decrease of the

activation enthalpy.

Substituent Effects. Kinetics of aryl-fluoride reductive

elimination from Pd(C₆H₄-p-X)F(Me₂NHC) models 1_OMe, 1,

1_NOshow substantial improvement with increasingly electron-

2

withdrawing para-substituents on the aryl ring (Figure 1).

Closely reminiscent of the experimental findings on related

C-S,^39a,bC-N,^35cand C-O^24c,freductive eliminations, com-

putational C-F data in Figure 1 show this to be primarily a

stabilizing TS effect (∆∆H^q) -5.0 kcal/mol), with analogous

(39) (a) Baran˜ano, D.; Hartwig, J. F. J. Am. Chem. Soc. 1995, 117, 2937-

2938. (b) Mann, G.; Baran˜ano, D.; Hartwig, J. F.; Rheingold, A. L.; Guzei,

I. A. J. Am. Chem. Soc. 1998, 120, 9205-9219. (c) Mann, G.; Hartwig, J.

F.; Driver, M. S.; Ferna´ndez-Rivas, C. J. Am. Chem. Soc. 1998, 120, 827-

828.

(40) Seppelt, K. Angew. Chem., Int. Ed. Engl. 1992, 31, 292-293.

9

J. AM. CHEM. SOC. VOL. 129, NO. 5, 2007 1347

A R T I C L E S

Yandulov and Tran

Figure 4. Computed HC(O)-F reductive elimination reactivity profile of

Pd(C(O)H)F(Me₂NHC).

and dative F^-(p)fAr(π*) interactions,⁴³promoted by para

electron-withdrawing substituents on Ar, comprise an important

contribution to the bonding in TS 1-2^q. Notably, structural

Figure 3. Computed effects of formation of hydrogen bonds between F

and external HF on Ph-F reductive elimination from Pd(Ph)F(Me₂NHC)L.

q

differences between 1-2_OMeand 1-2_NOaccompanying the

barrier (∆∆E^q) 9.5 kcal/mol) comparably and considerably.

This result is not surprising in view of the NBO fragment

charges also shown in Figure 3, which reveal a monotonic

decrease of the F^-value on going from 1 (-0.64) to 1-2^q

(-0.43) to 3 (-0.34 e^-). Thus, the fluoride becomes progres-

sively less “naked” during Ph-F reductive elimination, and

although TS 1-2^qdoes benefit from hydrogen bonding to F^-

in absolute terms, reactant 1 with a more basic F^-does so much

more, leading to a net increase of the activation barrier on going

2

-5.0 kcal/mol difference in their ∆H^qamount only to a 0.05 Å

q

2

shorter Pd‚‚‚Ar distance and a 2° sharper angle at F in 1-2_NO

.

Aspects of S_NAr have been thoroughly recognized in the

experimental mechanistic studies of related (aromatic) C-S,^39a,b

C-N,^35cand C-O^24c,freductive eliminations from Pd^II. For all

of these reaction categories, elimination kinetics were found to

correlate strongly with nucleophilicity of the heteroatom-

containing group,^35f,39cfurther corroborating the mechanistic

analogy based on aryl substituent effects. The importance of

the X group nucleophilicity to C-X reductive elimination

kinetics is striking, for example, in being responsible for

considerably faster rates of C-S bond-forming reactions in

comparison to the more thermodynamically favorable C-N and

C-O eliminations.^35fConversely, kinetic challenges facing C-F

reductive elimination in experimental practice may be attributed

to poor nucleophilicity of Pd^II-bound F^-svery different from

a “naked” F^-form⁴⁰sextrapolating the rationale put forth for

the kinetic order of heavy aryl-halide reductive eliminations from

Pd^II.^25cThus, it may appear as no surprise that hydrogen

bonding, reducing nucleophilicity of Pd-bound F^-in 1-2^q

further, is detrimental to the elimination kinetics (Figure 3).

However, this net kinetic effect can be attributed entirely to

greater relative stabilization of reactant 1 with its more basic

F^-. A substrate more susceptible to nucleophilic attack was

therefore considered to reveal more clearly any potential kinetic

benefits of promoting dative F^-fC interactions between

eliminating fragments in the TS. Indeed, Figure 4 shows how

dramatically more reactive a formyl group is compared to Ph,

to C(sp²)-F reductive elimination from the three-coordinate Pd^II

environment of 7. A computed ∆∆H^qreduction of 18.4 kcal/

mol on going from 1-2^qto 7-8^qis accompanied by a slight

decrease in reaction driving force (∆∆H° ) +0.4 kcal/mol)⁴⁴

with respect to the most stable Pd⁰adduct of the reductively

coupled R-F in each case, 3 and 9, and can thus be attributed

chiefly to the presence of a more accessible acceptor orbital

(π*_CdO) on the sp²carbon of R in 7. This comparison

demonstrates a clear advantage to the kinetics of C-F reductive

from 1 to 1_...HF

.

Promotion of Ar-F reductive elimination from PdArF(L) by

more weakly donating auxiliary L can therefore be safely

attributed to destabilization of oxidized 1, featuring largest

positive charge on Pd (Figure 3), relative to Pd⁰elimination

product 3 and less so TS 1-2^q. Recurrent throughout calcula-

tions (vide infra), this trend follows the textbook axiom of

reductive elimination being stimulated by increasing positive

charge at the metal,⁴¹such as by reducing auxiliary ligand donor

strength, or indeed their number (Figure 2). With partial positive

charge on the metal decreasing monotonically during the

reductive elimination (Figure 3), driving force is affected the

most by such auxiliary ligand variation, with a smaller promot-

ing influence also exerted on the elimination TS, producing

accord with the Hammond postulate. As the aryl ligand does

not remain auxiliary during elimination, thermodynamic effects

of aryl substituents are less transparent. However, clear structural

resemblance to a Meisenheimer intermediate born by TS 1-2^q

(in its cyclohexadienyl resonance form shown in Figure 1),

shows that any electronic effects of Ar substituents promoting

S_NAr reactivity⁷will be operative in Ar-F reductive elimination

as well. Therefore, significant stabilization of TS 1-2^qby para

q

2

electron-withdrawing groups on Ar (1-2_OMevs 1-2_NO

)

relative to both 1 and 3 can be at least partly attributed to an

S_NAr reaction character of the Ar-F reductive elimination (and

oxidative addition⁴²). Thus to some extent, F^-acts as a

nucleophile/nucleofuge with respect to the Ar in these reactions

(41) Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G. Principles and

Applications of Organotransition Metal Chemistry, 2nd ed.; University

Science Books: Mill Valley, CA, 1987; pp 322-333.

(43) Represented schematically in the cyclohexadienyl resonance form of TS

1-2^qin Figure 1 and illustrated in two canonical MO isosurfaces computed

for 1-2_NO^q(Supporting Information); for a recent computational analysis

2

(42) Oxidative addition of heavy aryl halides to [Pd⁰(PP)] in THF solVent was

computed to proceed via classical S_NAr mechanism followed by collapse

of the resulting ion pair: Senn, H. M.; Ziegler, T. Organometallics 2004,

23, 2980-2988.

of orbital interactions in transition states for C-X oxidative addition to

Pd(0)L_n, see: Ariafard, A.; Lin, Z. Organometallics 2006, 25, 4030-4033.

(44) This thermodynamic invariance is consistent with similarity of BDE’s¹⁰of

HC(O)-F (119) and Ph-F (126 kcal/mol).

9

1348 J. AM. CHEM. SOC. VOL. 129, NO. 5, 2007

Aryl-Fluoride Reductive Elimination from Pd(II)

A R T I C L E S

Figure 6. Computed effects of coordination of additional donor ligands

on Ph-F reductive elimination from three-coordinate PdPh(F)PR₃.

Figure 5. Computed reactivity profile of Pd(Ph)F(PMe₃) and related

analogs: Pd(C₆H₄-4-NO₂)F(P(t-Bu)₃) (numbers in cyan, right of center),

Pd(Ph)F(PPh₃) (numbers in blue, left of center), and Pd(C₆H₄-4-NO₂)F-

(PMe₃) (numbers in red, below center).

tions. Coordination of another PMe₃to 1 (∆H° ) -25.5, ∆G°

) -12.9 kcal/mol) is favorable by ∼12 kcal/mol more than

that to 10-11^qand 11, with the difference translating directly

into lowered driving force (∆H° ) +5.1 kcal/mol) and increased

activation barrier (∆H^q) 38.8 kcal/mol) for Ph-F reductive

elimination from 13 (Figure 6) compared to that from 10.

Although together with comparative reactivity of 5 and 5_η1

(Figure 2) this result speaks against the utility of four-coordinate

Pd^IIenvironment for mediating Ar-F reductive elimination,

elimination/oxidative addition of dative F^-(p)fC(π*) σ-sym-

metry interactions between eliminating fragments,⁴³a secondary

bonding mode geometrically constrained to be most favorable

in the TS of concerted processes. Calculations also suggest that

C-F reductive elimination from Pd^IIcan in principle operate

in a very fast kinetic regime (of synthetic applications), and

support the notion of ArC(O)-F reductive elimination from Pd^II

indeed mediating the carbonylative halide exchange fluorination

(fluorocarbonylation) of aryl halides (eq 1).^20,21

coordination of I^-to _PPh10 produces a notable exception.

3

Palladate 15 is computed to undergo Ph-F reductive elimination

with ∆H^q) 26.4 kcal/mol in the gas phase, a barrier of only

3. Ar-F Reductive Elimination from {Pd(Ar)F(PR3)}.

Figure 5 shows many of the trends computed for Ar-F reductive

elimination from Pd(Ar)F(NHC) in Figures 1-3 persisting in

the phosphine-supported Pd^IIenvironment. Thus, the PMe₃

analog of 1, Pd(Ph)F(PMe₃) (10), is predicted to be well suited

for reductive elimination of Ph-F (∆H^q) 25.1, ∆H° ) -5.6

kcal/mol), and even greater reactivity should be expected of its

three-coordinate PPh₃derivative (∆H^q) 24.1, ∆H° ) -7.7

kcal/mol). Consideration of the respective Ph-F elimination

∆H^q(∆H°) of 28.9 (+1.2) and 28.0 (-0.3) kcal/mol character-

izing such reactivity of 1 and _Cl21 suggests both NHC ligands

to be better donors to Pd^IIin 1 than PMe₃. The decreasing donor

2.3 kcal/mol higher than that computed for _PPh10 (Figure 6).⁴⁵

3

This finding is particularly encouraging in view of the ap-

preciable stability of the heavy halide analogs of 15, [Cat][Pd-

(Ph)X₂(PPh₃)] (X ) Cl, Br, I) relative to dimers [Pd(Ph)PPh₃(µ-

X)]₂and [Cat]X in CD₂Cl₂determined experimentally.^26e

However, it is also notable that analogous [Pd(Ar)(pyrrolyl)₂L]^-

palladates were found less reactive toward N-aryl pyrrole

reductive elimination than their neutral [Pd(Ar)(pyrrolyl)L₂]

derivatives.^39c

P(III)/Pd(II)fP(V)/Pd(0). Intramolecular oxidation of phos-

phines by Pd(II)/F^-has been identified experimentally as the

key process competing with Ar-F reductive elimination from

[Pd^II](Ar)F(PR₃) complexes (vide supra).^21b,26e,g,hAlthough

recent studies provided compelling evidence for formation of

fluorophosphines R₂PF by intramolecular R₂P-R/M-F ex-

change,^30c,dsupporting their proposed intermediacy in the overall

decomposition process,^26ecomplete mechanistic picture remains

tentative. To evaluate the feasibility of Ar-F reductive elimina-

tion from three-coordinate Pd(Ar)F(PR₃) in comparison to

intramolecular PR₃oxidation by Pd(II)/F^-, yet another decom-

position process was considered as a benchmark. Thermolysis

of Pd(Ph)F(PPh₃)₂showed the onset temperature of reversible

Ph-Ph/Pd-Ph exchange to be detectably lower than that of

irreversible PR₃oxidation by Pd(II)/F^-.^26eEvaluation of the

barrier to the former process, generally accepted to proceed via

reductive elimination of aryl-phosphonium ion and accelerated

ability in the auxiliary ligand series 1, _Cl21, 10, _PPh10 results

3

in lowering of activation enthalpy by the total of ∆∆H^q) -4.0

kcal/mol and proportionally greater increase in the driving force,

∆∆H° ) -8.9 kcal/mol, in accord with the Hammond postulate

and the basic arguments of decreasing positive charge at the

metal accompanying reductive elimination (vide supra). Also

consistent with Pd(Ar)F(NHC) results, substitution of the aryl

ring of 10 with electron-withdrawing p-NO₂promotes reductive

elimination with a sizable S_NAr-type kinetic effect (∆∆H^q)

-3.2, ∆∆H° ) -0.5 kcal/mol).

PdAr(F)L(PR₃). Increasing electrophilicity of Pd on going

from 1 to 10, while promoting Ph-F reductive elimination, also

stabilizes the fluoride-bridged dimer form [10]₂, by ∆∆H° )

-1.9 kcal/mol per monomer relative to 1f1/2[1]₂(Figures 1,

5). Like [1]₂, [10]₂would be unable to undergo Ph-F reductive

elimination via rapid pre-equilibrium breakup to 10 followed

by rate-limiting 10-11^q, with the added dimer dissociation

1/2∆G° raising the ∆G_obs^qfrom that of 10-11^qto 25.5 + 15.6

) 41.1 kcal/mol (cf. 43.3 for [10]₂; both are lower estimates

with dimer dissociation not being rate-limiting, vide infra), an

activation barrier again too high for practical synthetic applica-

(45) Activation enthalpies of both _PPh10-11^qand 15-16^qincrease by ∼3 kcal/

mol each in C₆H₆with net, fav³orable solvation energy being ∼20 kcal/

mol greater for the anionic derivatives.

9

J. AM. CHEM. SOC. VOL. 129, NO. 5, 2007 1349

A R T I C L E S

Yandulov and Tran

by ligand dissociation from square-planar precursors,⁴⁶could

thus serve as an approximate (lower bound) measure of the

R₃P-Pd(II)/F^-oxidation kinetics.

Scheme 1. Synthesis and Solution Characterization of 17 and 18

Figure 5 shows migration of Ph from Pd in 10 to coordinated

PMe₃to face a kinetic barrier (10-12^q, ∆H^q) 21.1 kcal/mol,

see also Supporting Information) considerably lower than that

of reductive elimination of Ph-F (10-11^q, ∆H^q) 25.1 kcal/

mol). The Ph-PMe₃coupled product 12, decidedly unstable

with respect to 10 (∆H° ) +16.2 kcal/mol), adopts a gas-phase

structure best described by a combination of resonance forms

of 1,2η-C₆H₅-PMe₃⁺adduct of Pd⁰(F^-) and Pd^IIphosphoranide

Pd(PPhMe₃)F, featuring short Pd-P (2.31 Å, cf. 2.25 Å in 10)

and Pd‚‚‚C1,C2 (2.10, 2.30 Å) distances.³⁸Presence of electron-

withdrawing p-NO₂on the migrating phenyl group in 10-12^q

raises the barrier by ∆∆H^q) 2.9 kcal/mol, consistent with

experimental trends,^46dand stabilizes the Ar-PMe₃coupled

structure by ∆∆H° ) 0.9 kcal/mol. As a result of the opposite

substituent effect p-NO₂group has on the Ph-F elimination

kinetics, latter reaction becomes more facile than PMe₃arylation

(∆∆H^q) -2.1 kcal/mol) for 10_NO. All changes resulting from

2

p-NO₂substitution in 10 duplicate substituent effects in 1 (Figure

1). Thus, subject to the validity of the above assumption of the

relative activation barriers of P-C and P-F bond-forming

decomposition processes in a three-coordinate [Pd^II](Ar)F(PR₃)

environment, computational results in Figure 5 suggest that

Ar-F reductive elimination may be able to compete with

irreversible oxidation of PR₃by Pd(II)/F^-, with improved

selectivity for C-F bond formation displayed by more electron-

deficient Ar substrates.

NO₂)L(µ-F)]₂featuring sterically demanding L ) P(o-Tolyl)₃

(17) and P(t-Bu)₃(18) were prepared starting from the known⁴⁷

[Pd(C₆H₄-4-NO₂)P(o-Tolyl)₃(µ-I)]₂(Scheme 1), isolated and

fully characterized. We found AgF metathesis²¹of the Pd iodide

precursor to require sonication over simple stirring,³²but also

a catalytic amount of P(o-Tolyl)₃. Both 17 and 18 maintain

fluoride-bridged dimer structures in solution, even at elevated

temperatures (vide infra). Stability of dimeric [PdArL(µ-F)]₂

ground states with L of the largest cone angles characterized

so far in a {Pd}F coordination environment contrasts with the

structural preferences of the X ) Br and I analogs of P(t-Bu)₃

derivative 18²⁵, although does follow the trend of related Pd-

(Ar)X(Q-Phos) that adopt dimeric structure with X ) Cl but

not Br or I.^47b

Steric Factors. The key advantage of the [Pd^II](Ar)F(PR₃)

environment in practice over that based on NHC’s is the

availability of a series of sterically demanding PR₃ligands,

whose bulk may serve to additionally facilitate C-F reductive

elimination process. The series of Pd(Ar)X(P(t-Bu)₃) complexes

investigated by Hartwig, in particular, exemplify such remark-

able effects of bulky and strongly electron-donating phosphines,

both in the stability of the unique 14-electron complexes with

weak agostic interactions and their unprecedented reactivity by

reductive elimination of heavy aryl halides.^25,35eComputational

data in Figure 5 reveal the magnitude of steric effects on

elimination barrier that might be accessible in the case of Pd-

(Ar)F(P(t-Bu)₃): reductive elimination of (O₂N-4-C₆H₄)-F from

_PtBu10_NO, evaluated with ONIOM models, requires ∆H^q3.9

Solution Structures. Variable-temperature NMR studies of

17 and 18 in several solvents revealed existence of intercon-

verting syn and anti isomeric forms of the dimers and allowed

measurement of individual spectroscopic data, while iterative

line shape analysis⁴⁸enabled determination of thermodynamic

and activation parameters governing the antiTsyn equilibrium

in each complex. Scheme 1 summarizes these data, and Figure

7 shows representative of the observed and fitted NMR

lineshapes (complete results can be found in Supporting

Information).

3

2

kcal/mol lower than that from its PMe₃analog 10_NOand is

2

subject to a 6.0 kcal/mol greater driving force, although each

change is likely overestimated in comparison to experiment due

to poor representation of CH‚‚‚Pd agostic interactions in both

q

_PtBu10_NOand _PtBu10-11_NO(vide supra).²⁹

3

2

3

2

Collectively, computational studies show that auxiliary ligands

of high steric demand but moderate donor strength, together

with electron-poor aryl substrates, create the most promising

environment for Ar-F reductive elimination from three-

coordinate Pd^II.

B. Experimental Studies. 1. Synthesis of [Pd(C₆H₄-4-

NO₂)L(µ-F)]₂, L ) P(o-Tolyl)₃, P(t-Bu)₃. Following the insight

gained from computational studies, the novel [Pd(C₆H₄-4-

For both 17 and 18 isomerization of anti to syn form is

marginally exothermic (∆H° ) -1 kcal/mol), with antiTsyn

equilibrium ∆S° sufficiently negative (-5 cal/(mol K)) to render

syn form the minor isomer in all but one spectra obtained,

diminishing in equilibrium concentration with increasing tem-

perature. Increasing solvent polarity shifts antiTsyn equilibrium

(46) (a) Kong, K.-C.; Cheng, C.-H. J. Am. Chem. Soc. 1991, 113, 6313-6315.

(b) Segelstein, B. E.; Butler, T. W.; Chenard, B. L. J. Org. Chem. 1995,

60, 12-13. (c) Morita, D. K.; Stille, J. K.; Norton, J. R. J. Am. Chem. Soc.

1995, 117, 8576-8581. (d) Goodson, F. E.; Wallow, T. I.; Novak, B. M.

J. Am. Chem. Soc. 1997, 119, 12441.

(47) (a) Paul, F.; Patt, J.; Hartwig, J. F. Organometallics 1995, 14, 3030-3039.

(b) Barrios-Landeros, F.; Hartwig, J. F. J. Am. Chem. Soc. 2005, 127, 6944-

6945.

(48) Budzelaar, P. H. M. gNMR, version 5.0.6.0; IvorySoft: Englewood, CO,

2006.

9

1350 J. AM. CHEM. SOC. VOL. 129, NO. 5, 2007

Aryl-Fluoride Reductive Elimination from Pd(II)

A R T I C L E S

Scheme 2. Solution Dynamics Exhibited by 17 and 18

both spectra exhibiting notable broadening already at 22 °C.

Retention of J_PFcoupling, at its population-weighed exchange-

averaged value, in the high-temperature limit (Figure 7) identi-

fies this dynamic process as the antiTsyn-18 interconversion.

Rate constants in several solvents obtained from iterative line

shape fitting yielded, from Eyring analysis (Supporting Informa-

tion), ∆H_synTanti-18^q) 14 kcal/mol and ∆S_synTanti-18^q) -4 cal/

(mol K), latter being expectedly small for an intramolecular

process (Scheme 1). The free energy barrier to antiTsyn

exchange, ∆G_+60°^q, in 18 is lower than that in 17 by 1.6(3)

kcal/mol. The minimal physical process manifested in line-

broadening and coalescence of NMR signals of syn- and anti-

isomers of 17 and 18 and whose activation parameters are

determined from the Eyring plots corresponds to cleavage of a

single Pd-F bond (a, a′, ...) necessarily combined with

isomerization of the resulting T-shaped intermediate (b, Scheme

2), which serves to exchange A/A′ and X/X′ sites. Because

isomerization of T-shaped d⁸ML₃complexes is facile in the

absence of strong driving force effects,⁵⁰the obtained activation

parameters (∆H_synTanti^q, ∆G_synTanti^q) can be taken as close upper

estimates on the values corresponding to single Pd-F bridge

cleavage.⁵¹

Initial VT NMR experiments in which samples of 18 were

maintained at elevated temperatures for extended periods of time

in fact showed complete loss of J_PFcoupling at +80 °C in PhF

and +95 °C in C₆D₆(Figure 7A). As shown in the Supporting

Information, spectra of these thermolyzed samples retained some

degree of line-broadening on cooling back to 22 °C. However,

addition of 10 equiv of CsF to these samples restored the line

shape observed originally at 22 °C and further enabled direct

observation of the exchange-averaged J_PFin the ¹⁹F and ³¹P-

{¹H} spectra at the +95 °C temperature limit, recorded

immediately after the minimum time required for temperature

equilibration of the samples in the pre-heated NMR probe. Given

that accompanying high-T °C ¹H NMR spectra featured a

single J_PHcoupling in one of the two signals of C₆H₄-4-NO₂in

the absence or presence of CsF, including spectra of 17 at the

high-T °C limit of +80, {Ar-Pd-L} unit remains stoichiomet-

rically intact on the NMR time scale (τ>0.2 s) up to +95 °C in

18 and +80 °C in 17. The loss of J_PFcoupling in 18 and its

reestablishment on treatment with CsF can therefore be at-

tributed to formation of trace quantities of HF from decomposi-

Figure 7. Variable-temperature ¹⁹F and ³¹P{¹H} spectra of 18 in PhF and

C₆D₆showing the effects of dynamic synTanti equilibrium. Iteratively fitted

line shape is overlaid in red.

toward syn isomers, possessing greater dipole moments, with

K_antiTsynat 22 °C equal to 0.63 (C₆D₆), 1.18 (CD₂Cl₂) for 17

and 0.14 (C₆D₆), 0.41 (PhF), 0.46 (CD₂Cl₂) for 18. Hindered

rotation of P(o-Tolyl)₃ligands⁴⁹in 17 complicates the observed

line-broadening effects and restricts the amount of spectroscopic

data available for anti- and syn-17 isomers to those shown in

Scheme 1. Observation of ordinary overlapping doublets in ³¹P-

{¹H} spectrum of 17 at 22 °C limits J_FFcoupling in anti-17 to

e60 Hz, so as to conceal second-order effects in the line width.

As the temperature is raised, spectra of 17 undergo sharpening,

due to accelerated rotation of P(o-Tolyl)₃, followed by broaden-

ing resulting from antiTsyn-17 interconversion; latter is most

clearly indicated by the coalescence of o-Me signals of

1

respective isomers in H NMR spectra. Line shape analysis of

60 °C spectra yielded an exchange rate constant of 6 × 10¹s^-1

q

that corresponds to ∆G_60°) 17 kcal/mol.

Complex 18 does not suffer complications from hindered

rotation of P(t-Bu)₃in the examined temperature range, which

enabled a more extensive solution characterization. Low-

temperature ¹⁹F and ³¹P{¹H} NMR spectra clearly revealed the

expected AA′XX′ patterns for anti-18, overlapped by ABX₂

patterns of syn-18. Both syn- and anti-18 feature J_FFcoupling

constants (103, 120 Hz) substantially larger than those reported

for PCy₃and PⁱPr₃(60, 63 Hz) analogs^26gof syn-18. Raising

the temperature led to broadening and eventual coalescence of

both ³¹P{¹H} and ¹⁹F NMR signals of syn- and anti-18, with

(50) (a) Thorn, D. L.; Hoffmann, R. J. Am. Chem. Soc. 1978, 100, 2079-2090.

(b) Romeo, R.; Minniti, D.; Lanza, S. Inorg. Chem. 1979, 18, 2362-2368.

(c) Komiya, S.; Albright, T. A.; Hoffmann, R.; Kochi, J. K. J. Am. Chem.

Soc. 1976, 98, 7255-7265. (d) Kayaki, Y.; Tsukamoto, H.; Kaneko, M.;

Shimizu, I.; Yamamoto, A.; Tachikawa, M.; Nakajima, T. J. Organomet.

Chem. 2001, 622, 199-209.

(51) It appears reasonable to expect bridge cleavage leading to the formation

of a T-shaped PdAr(µ-F)L fragment with L trans to the vacant site to

proceed in a concerted manner with its isomerization to the more sterically

stable geometry with Ar trans to the vacant site, i.e. that processes a′ and

b in Scheme 2 and the like occur together in a single elementary step.

(49) Notheis, J. U.; Heyn, R. H.; Caulton, K. G. Inorg. Chim. Acta. 1995, 229,

187-193.

9

J. AM. CHEM. SOC. VOL. 129, NO. 5, 2007 1351

A R T I C L E S

Yandulov and Tran

positional disorder. One of the PhF molecules participates in a

peculiar π-stacking interaction with one of the C₆H₄-4-NO₂

ligands (PhF‚‚‚C(13)C₅H₄-4-NO₂) 3.52 Å), protruding in the

cleft between the syn Ar rings to the extent compatible with

the (PhF)H‚‚‚F(2) hydrogen bond (H‚‚‚F 2.53 Å);⁵³the same

PhF molecule forms another (PhF)H‚‚‚F(2) hydrogen bond

(H‚‚‚F 2.65 Å) to the F(2) site in another dimer. An intermo-

lecular (P(t-Bu))CH‚‚‚F(2) ) 2.92 Å interaction brings the

number of H‚‚‚F e 3.0 Å contacts to F(2) to three, whereas

F(1) forms six such interactions (1.98-2.84 Å), all intermo-

lecular with P(t-Bu)₃hydrogens. Extensive participation of one

fluoride of syn-18 in hydrogen bonding with surrounding P(t-

Bu)₃C-H bonds^26gbears direct implications for its thermal

reactivity (vide infra). Severe steric repulsion between P(t-Bu)₃

and cis-aryl ligands is evident in the bending of aryl rings at

ipso-C13, C31 by 13 and 14°, respectively, from coplanarity

with Pd-C_ipsobonds, as well as widening of the (t-Bu)-P-Pd

angles to the (C9)Me₃and (C23)Me₃groups to 120°. Despite

the notable distortions, five short CH‚‚‚C(Ar) contacts remain

between the P(t-Bu)₃hydrogens and C₆H₄-4-NO₂carbons,

ranging from 2.31 to 2.54 Å. Of the total of 12 structures with

P(t-Bu)₃bound in a square-planar coordination environment

available from CSD,⁵⁴the only one with a cis-aryl ligand⁵⁵

shows similar bending of the Ph group from coplanarity with

Pd-C_ipso, by 8°. Pd-F bond distances (2.078(4)‚‚‚2.140(4) Å)

in syn-18 compare well with those in the other two structurally

characterized fluoride-bridged dimeric Pd complexes (2.0984-

(8)‚‚‚2.1336(11) Å).^26g

Figure 8. ORTEP drawing (50% probability ellipsoids) of the crystal

structure of 18 showing the intercalated molecule of PhF. All heavy atoms

were refined anisotropically. Selected geometrical parameters (Å, deg): Pd-

(1)-C(13), 1.944(7); Pd(2)-C(31), 1.929(7); Pd(1)-F(2), 2.080(4);

Pd(2)-F(2), 2.078(4); Pd(1)-F(1), 2.140(4); Pd(2)-F(1), 2.120(4); Pd-

(1)-P(1), 2.283(2); Pd(2)-P(2), 2.279(2); F(2)-Pd(1)-F(1), 76.00(15);

F(2)-Pd(2)-F(1), 76.48(15); C(13)-Pd(1)-P(1), 97.3(2); C(31)-Pd(2)-

P(2), 98.5(2); F(1)-Pd(1)-P(1), 102.39(12); F(1)-Pd(2)-P(2), 100.50-

(12); Pd(2)-F(1)-Pd(1), 101.84(16); Pd(2)-F(2)-Pd(1), 105.39(17).

Scheme 3. Proposed Intermolecular F^-Exchange in 18 Mediated

by HF

2. Thermal Reactivity of [Pd(C₆H₄-4-NO₂)L(µ-F)]₂. L )

P(o-Tolyl)₃, 17. Thermolysis of 17 in C₆D₆at 60 °C required

near a week for complete conversion and resulted in a mixture

of reductively coupled and protolyzed aromatic compounds

originating from both ArdC₆H₄-4-NO₂and o-Tolyl substituents

of the phosphine (Scheme 4).⁵⁶PdL₂^47aformed and decayed as

the only identifiable intermediate. In addition to large quantities

of Pd⁰and free L ) P(o-Tolyl)₃, the resulting mixture contained

57a

a minor amount of PF₂(o-Tolyl)₃(∼10% in F, ¹⁹F NMR),

whereas GC-MS and ESI-MS also indicated presence of P(o-

Tolyl)₂Ar, PF₂(o-Tolyl)₂Ar and complete absence of phospho-

nium P(o-Tolyl)₃Ar⁺(after redissolution in CDCl₃) or the

desired Ar-F. Qualitatively, the observed organic products

signify two key processes dominating thermal reactivity of 17

(Scheme 4): (p) F/Ar ligand redistribution⁵⁸followed by

reductive elimination of Ar-Ar from the Pd(Ar)₂L fragment,

and (q) Pd-Ar/P-(o-Tolyl) aryl group exchange^46,26eproducing

(o-Tolyl)-analogs of {Pd}Ar intermediates and leading to

tion of 18 on thermolysis (vide infra). One mechanism by which

J_PFcoupling in 18 could be lost with help of HF is the

intermolecular F^-exchange, possibly proceeding with

minimal disruption of F^--bridged dimer structure as shown in

Scheme 3.

These high-temperature NMR studies thus offer no conclusive

evidence for complete dimer cleavage. Broadening of the

limiting 63 Hz binomial triplet in the 95 °C ³¹P{¹H} spectrum

(Figure 7A), though likely due to traces of HF, limits the lifetime

of the two monomers connected via at least one bridge to at

least ∼1/(50 Hz) or τ > 20 ms and complete dimer dissociation

(53) C-H bonds were extended to 1.083 Å for all reported H‚‚‚X measurements.

(54) The Cambridge Structural Database, version. 5.27, November 2005; The

Cambridge Crystallographic Data Centre: Cambridge, U.K., 2006.

(55) Liu, X.; Hartwig, J. F. Org. Lett. 2003, 5, 1915-1918.

(56) All yields are given in Pd equivalents, i.e. 1 mol of Ar-Ar formed per 1

mol of Pd dimer corresponds to 100%. Products quantified in Schemes 4

and 5 were identified directly in the reaction mixtures by NMR and

quantified by integration vs internal standard (( 5%). Products identified

by GC-MS only may have originated from additional thermal reactivity at

T °C of up to 280 during GC analysis. See Supporting Information for

additional details.

q

∆G_+95°C> 19 kcal/mol.

X-ray Structure of syn-18×(PhF)₂. Dimeric nature of 18

was substantiated further by a crystal structure determination

(Figure 8).⁵²18 crystallizes from PhF at -35 °C as the minor,

syn-isomer with two PhF molecules per dimer that form separate

extended hydrogen-bonded networks and suffer from minor

(57) (a) Holmes, R. R.; Holmes, J. M.; Day, R. O.; Swamy, K. C. K.;

Chandrasekhar, V. Phosphorus, Sulfur Silicon Relat. Elem. 1995, 103, 153-

169. (b) Herrmann, W. A.; Brossmer, C.; Reisinger, C.-P.; Riermeier, T.

H.; O¨ fele, K.; Beller, M. Chem.-Eur. J. 1997, 3, 1357-1364. (c) Crozet,

M. P.; Vanelle, P. Tetrahedron Lett. 1985, 26, 323-326. (d) Fild, M.;

Schmutzler, R. J. Chem. Soc., A 1970, 2359-2364. (e) Dura`-Vila`, V.;

Mingos, D. M. P.; Vilar, R.; White, A. J. P.; Williams, D. J. J. Organomet.

Chem. 2000, 600, 198-205.

(52) Crystallographic details for [Pd(C₆H₄-4-NO₂)P(t-Bu)₃(µ-F)]₂× (PhF)₂(18

× (PhF)₂), C₄₈H₇₂F₄N₂O₄P₂Pd₂, 1091.82 g/mol, Monoclinic, P2₁/n, a )

15.473(4), b ) 12.450(3), c ) 26.373(7), â ) 100.704(4)°, V ) 4992(2)

Å³, T ) -99 °C, Z ) 4; R1 ) 0.0450, wR2 ) 0.0901 ([I > 2σ(I)]); R1 )

0.0857, wR2 ) 0.1028 (all data); GOF ) 1.018.

(58) Ca´rdenas, D. J.; Mart´ın-Matute, B.; Echavarren, A. M. J. Am. Chem. Soc.

2006, 128, 5033-5040.

9

1352 J. AM. CHEM. SOC. VOL. 129, NO. 5, 2007

Aryl-Fluoride Reductive Elimination from Pd(II)

A R T I C L E S

Scheme 4. Thermal Reactivity of 17: Identified Products with Their Yields (Left) and Proposed Reaction Pathways (Right)

Scheme 5. Thermal Reactivity of 18: Identified Products with Their Yields (Left) and Proposed Reaction Pathways (Right)

corresponding (o-Tolyl)-derived organic products. The Pd(II)

coproduct of (p), {PdF₂L}, appears to react by metalation

forming analogs of (poorly soluble) Herrmann’s dimers,^57brather

than F^--mediated P(III)/Pd(II)fP(V)/Pd(0) intramolecular

redox^{21b,26e,g,h,30}judging from the low yield of PF₂(o-Tolyl)₃

alone⁵⁹and the lack of any other appreciable ¹⁹F signals.⁶⁰The

HF released during o-Tolyl metalation at this, or possibly any

other {Pd}F(P(o-Tolyl)₃) stage, serves as a proton source in the

formation of Ar-H and (o-Tolyl)H. o-Tolyl metalation would

also be consistent with the observed HTD exchange with C₆D₆

solvent, proceeding to the extent of ∼4.5 hydrogens per Pd, at

least as a means to introduce reactive H into the system. If the

o-Tolyl groups of the observed organic products are considered

equivalent to the original Ar, exchanged via (q), the Ar balance

observed in the 60 °C reaction is 77%. At 80 °C, thermolysis

of 17 requires under 60 h to completion and yields the same

products with both a decreased proportion of Ar/o-Tolyl

exchanged products and an improved Ar balance (100 ( 10%).

Presence of iodide ions did not aid in the formation of the

Ar-F reductive elimination product, as was hoped on the basis

of computational results for 15 in Figure 6. 17 exchanges both

fluorides for I^-with added 5 equiv. of [Bu₄N]I over an hour at

RT in C₆D₆and begins to react via pathway p (Scheme 4) with

formation of biaryl to an appreciable extent. A 3-h thermolysis

at 60 °C completes conversion of Ar into Ar-Ar (73%), Ar-

exchange between 17 and Bu₄NI, neither free 1-butene nor Bu₃N

were observed in the reaction mixture, possibly participating

in coordination equilibria with the reduced Pd products formed

concomitantly.

L ) P(t-Bu)₃, 18. Thermolysis of 18 in C₆D₆requires similar

time to completion as does 17, but yields a more complex

mixture of products (Scheme 5). At 60 °C, isobutylene is the

major organic product (g31% measured by integration of the

dissolved portion), formed together with Ar-H (22%), Ar-Ar

(8%) and the Heck arylation product of isobutylene, ArCHd

CMe₂^57c(6%).⁵⁶Pd is accounted for with PdL₂(14%) in addition

to ample Pd⁰, while P(t-Bu)₃gives rise to P(O)F(t-Bu)₂(15%)

and PF₂Ar(t-Bu)₂(3%). With identity of P(O)F(t-Bu)₂assured

by EI-MS and close agreement of NMR data with literature,^57d

that of the novel PF₂Ar(t-Bu)₂was established on the basis of

closer agreement of its observed spectroscopic data with those

57a

of PF₂Ph(t-Bu)₂

than PF₂(t-Bu)₃,^{57d 19}F NMR pattern

a

1

featuring coupling (J_FH) 2.8 Hz, from H NMR) to 18 rather

than 27 equiv protons and a fully assigned ¹H spectrum

(Supporting Information); HRMS was additionally used to verify

the elemental composition of P(O)Ar(t-Bu)₂observed in ESI-

MS and GC-MS of the reaction mixture. The predominant

reactivity mode of 18 thus appears to be γ-hydrogen elimination

from coordinated P(t-Bu)₃(r), with bridging fluoride functioning

as a base (Scheme 5). Spatial proximity of numerous P(t-Bu)₃

hydrogens to F, revealed in the crystal structure of 18, likely

assists the initial proton-transfer event possibly leading to a

cyclometalated {Pd}(ηC,P-CH₂CMe₂P(t-Bu)₂) intermediate

prior to isobutylene elimination. Such metalation followed by

isobutylene elimination was thoroughly documented as part of

thermal reactivity of Ni(COD)₂/(t-Bu)₂NHC system.⁶¹Palladium

phosphide {Pd^II}P(t-Bu)₂intermediate resulting from isobuty-

2-

(o-Tolyl) (18%) and (o-Tolyl)₂(7%), with HF₂^-(70%), SiF₆

(16%) and PF₂(o-Tolyl)₃(7%) providing a solid mass balance

for the total F. Although the latter distribution is indicative of

Hoffman elimination of [Bu₄N]F, produced from the halide

(59) Although ESI-MS of the reaction mixture, after redissolution in CDCl₃,

did show metalated Pd(CH₂-C₆H₄-2-P(o-Tolyl)₂)⁺and Pd(CH₂-C₆H₄-2-

P(o-Tolyl)₂)(P(o-Tolyl)₃)⁺fragments, their presence in ESI-MS spectrum

is typical of many Pd(P(o-Tolyl)₃) complexes, i.e. metalation takes place

readily during ESI.

(60) A reviewer suggested PF₂(o-Tolyl)₃to react with low-ligated, highly reactive

Pd(0) species instead.

(61) Caddick, S.; Cloke, F. G. N.; Hitchcock, P. B.; Lewis, A. K. de K. Angew.

Chem., Int. Ed. 2004, 43, 5824-5827.

9

J. AM. CHEM. SOC. VOL. 129, NO. 5, 2007 1353

A R T I C L E S

Yandulov and Tran

Figure 9. Computed structure of syn-18 and reactivity profiles of [Pd(C₆H₄-4-NO₂)(L)(µ-F)]₂(L ) P(t-Bu)₃, [_PtBu10_NO]₂; L ) PMe₃, [10_NO]₂, in blue).

3

2

lene/HF elimination in the case of 18 could ostensibly undergo

1,2-migration^30dof Ar from Pd to furnish the modified phos-

phine P(C₆H₄-4-NO₂)(t-Bu)₂, or that of a fluoride to yield PF-

(t-Bu)₂, traces of which were identified by GC-MS. The novel

PAr(t-Bu)₂thus formed can be correlated to the observed PF₂-

Ar(t-Bu)₂via F^--mediated P(III)/Pd(II)fP(V)/Pd(0) intramo-

lecular redox,^{21b,26e,g,h,30}with hydrolysis of PF₂Ar(t-Bu)₂or-

{Pd^II}P(t-Bu)₂by adventitious moisture possibly responsible for

the formation of P(O)F(t-Bu)₂. Repeating the reaction at 80 °C

for 60 h increased the PF₂Ar(t-Bu)₂: P(O)F(t-Bu)₂ratio from

1:5 to 6:4 of the same total. Containment of the reaction mixture

in a Teflon NMR tube liner produced nearly identical results,

yet in the presence of 10 equiv. of CsF in an all-glass NMR

tube at 80 °C formation of the oxofluorophosphorane was

strongly suppressed, PF₂Ar(t-Bu)₂:P(O)F(t-Bu)₂) 9:1. These

results suggest adventitious water associated with solid sample

of 18, rather than that from glass etching with HF and adsorbed

by added CsF, as the source of P(O)F(t-Bu)₂oxygen. Notably,

volatiles vacuum-transferred from the reaction mixture contained

no trace of isobutane, which necessitates hydrolysis of PF₂Ar-

(t-Bu)₂to be selective for the formation of H-Ar or, more likely,

that P-O bond formation takes place at the {Pd^II}P(t-Bu)₂

phosphido stage in a reaction parallel and competing with

formation of P(C₆H₄-4-NO₂)(t-Bu)₂. Formation of the other key

products, Ar-Ar and Ar-H, can be envisioned to proceed in

the same manner as in the case of 17, with the added possibility

of PAr(t-Bu)₂supporting Pd during the F/Ar ligand redistribution

(p). Still, the observed products summarized in Scheme 5

amount to a mass balance of Ar of only 38% in the 60 °C

reaction, which increases to 48% in the 80 °C/60 h run. Other

products of thermolysis of 18, featured in complex and

anticipated Ar-F reductive elimination from the palladate

analog of 15 (Figure 6).

Computational Analysis of [_PtBu10_NO]₂Reactivity. The

3

2

failure of Pd(C₆H₄-4-NO₂)F(P(t-Bu)₃) to produce Ar-F by

reductive elimination to a significant extent, as was expected

on the basis of computational results, was predetermined by

the exceeding stability of the fluoride-bridged dimer 18 to

complete dissociation into monomers. High-temperature NMR

studies (vide supra) produced activation parameters for intramo-

lecular rearrangement involving cleavage of one bridge at a time,

but gave no conclusive evidence with regard to kinetics of

complete dimer breakup at +95 °C other than that it remains

slow on the NMR time scale (τ > 20 ms, ∆G_+95°C^q> 19 kcal/

mol). Calculations provided this insight, with faithful ONIOM

models of 18, [_PtBu10_NO]₂(Figure 9) reproducing quite well

3

2

both the key geometric parameters observed in the crystal

structure of syn-18, including bending of the aryl rings and t-Bu

groups and multiple short H‚‚‚F contacts between P(t-Bu)₃

hydrogens and F(1), as well as the near ergoneutrality of the

syn and anti-isomers. If the activation free energy for dissocia-

tion of [_PtBu10_NO]₂into two molecules of _PtBu10_NO, ∆G^q(-

3

2

3

2

[

_PtBu10_NO]_2f1), is approximated by the dissociation ∆H°, which

3 2

is equivalent to ∆G^qfor dimerization being primarily entropic

in origin, kinetics of Ar-F elimination from [_PtBu10_NO]₂via

3

2

the two-step process depicted in Figure 9 can be evaluated

explicitly. Steady-state rate expression for the loss of [_PtBu10_NO

]

2

3

is bound by two limiting forms, a first-order dependence with

rate-limiting dimer dissociation step and a half-order rate

dependence on concentration of [_PtBu10_NO]₂, in which _PtBu10_NO

2

3

2

3

is formed in a rapid pre-equilibrium and undergoes reductive

elimination as the rate-determining step.⁶²The activation free

energy, observed in at least the initial stages, is the higher of

1

numerous aromatic H NMR patterns remain to be identified.

∆G₁) ∆G^q([_PtBu10_NO

]

_2f1) and ∆G_1/2) [1/2 ∆G°(-

q

Most importantly however, a trace of the desired Ar-F

(∼0.05%) could be identified with certainty by EI-MS in the

GC-MS of several 80 °C reaction mixtures.⁵⁶

3

2

[

10

]

_2f1) + ∆G^q(

10-11 ^q) + RTln2], for the

PtBu₃NO₂

PtBu₃

NO₂

limiting first- and half-order rate dependence respectively.

Adoption of either limit is subject to the magnitude of the term

18 shows no appreciable reactivity with 4 equiv of [Bu₄N]I

in C₆D₆over several hours at RT but cleanly yields Ar-Ar

(92.2%), HF₂^-(76.2%), and SiF₆^2-(26.2%) after 3 h at 60 °C.

a ) 16k₁k_-1/k₂) 4 exp(-2[∆G₁- ∆G_1/2^q]/RT), with a ,

1/[18] and a . 1/[18] producing first- and half-order kinetics,

respectively. Thus, ONIOM data in Figure 9 predict reductive

elimination of Ar-F from 18 to be rate-limited by the initial

2

q

1

Ample free 1-butene and Bu₃N are seen by H NMR of the

57e

product mixture, with PdL₂and [Pd^IL(µ-I)]₂accounting for

dimer dissociation (a × [18] ) 0.0001 at 80 °C and 20 mM

all signals in ³¹P{¹H} spectrum together with a minor peak at

91.3 ppm. These observations show 18 to undergo halide

exchange of Pd-F with added I^-, followed by I^-/Ar ligand

redistribution (p) in PdAr(I)P(t-Bu)₃and reductive elimination

of Ar-Ar, and Hoffman elimination of [Bu₄N]F, instead of the

initial[18]), with ∆G_obs) ∆G^q([_PtBu10_NO

]

2

_2f1) ≈ ∆H°(-

q

3

[

_PtBu10_NO]_2f1) ) 27 kcal/mol. Partial cleavage of 18, whose

3 2

(62) Espenson, J. H. Chemical Kinetics and Reaction Mechanisms, 2nd ed.;

McGraw-Hill: New York, 1995; p 82.

9

1354 J. AM. CHEM. SOC. VOL. 129, NO. 5, 2007

Aryl-Fluoride Reductive Elimination from Pd(II)

A R T I C L E S

Table 2. Correlations between Activation and Equilibrium

∆H° e ∆H_synTanti-18^q) 14 kcal/mol (Scheme 1), could offer a

more facile pathway to the overall Ar-F reductive elimination

only if the (unobserved) singly bridged dimer is unstable by

∆H° < 9 kcal/mol relative to 18, because reductive elimination

from either of the dimer parts connected via a single F^-bridge

would require an additional ∆H^q> ∆H^q(_PtBu10-11_NO^q) ) 18

Enthalpies (kcal/mol) of Ar-F Reductive Elimination Computed for

Various L_nPd^II(Ar)F Complexes

L_nPd^II(Ar)F

L_n

Ar

∆

H^q

∆H°

R

^a(R²)

1

_Cl21

15

10

_PPh10

Me₂NHC

Ph

28.9

1.2

0.50

(0.92)

Cl₂Me₂NHC

(I^-)PPh₃

PMe₃

28.0 -0.3

26.4 -1.0

25.1 -5.6

24.1 -7.7

3

2

kcal/mol (Figure 5).⁶³The observed reaction time scale, 60 h

at 80°, implies an effective t_1/2≈ 10 h and ∆G_obs^qof 28.4 kcal/

mol, which is close enough to the ONIOM estimate of 27 kcal/

mol to suggest that the true reductive elimination barrier from

18 is not too far above that for C-H activation of P(t-Bu)₃, the

predominant reactivity observed, yet high enough to afford only

marginal traces of the Ar-F.⁶⁴

PPh₃

3

5

1_...HF

1

5_η1

13

10

η²-(MeNHC)₂CH₂Ph

50.6

39.7

28.9

28.4

38.8

22.4

13.6

1.2

0.5

5.1

Me₂NHC/F...HF

Me₂NHC

Ph

η¹-(MeNHC)₂CH₂Ph

0.94

(0.96)

(PMe₃)₂

PMe₃

Ph

25.1 -5.6

10_NO

PMe₃

C₆H₄-4-NO₂21.9 -6.1

C₆H₄-4-NO₂18.0 -12.1

Computational data for the PMe₃analogs in Figure 9 put the

effect of P(t-Bu)₃in perspective. Intramolecular steric repulsion

2

_PtBu10_NOP(t-Bu)₃

3

2

with P(t-Bu)₃ligands makes dissociation of [_PtBu10_NO]₂20 kcal/

3

2

1_OMe

1

Me₂NHC

C₆H₄-4-OMe 30.5

Ph 28.9

C₆H₄-4-NO₂25.5 -0.2

1.8

1.2

2.4

(1.00)

mol (per dimer) more favorable in ∆H° than that of [10_NO]₂,

2

in addition to lowering the ∆H^qof Ar-F reductive elimination

by 3.9 kcal/mol. With the dimer dissociation predicted to be

rate-limiting in the overall reductive elimination starting from

1_NO

2

10

10_NO

PMe₃

Ph

25.1 -5.6

21.9 -6.1

7.0

C₆H₄-4-NO₂

2

1

7

Me₂NHC

Ph-F

HC(O)-F

28.9

10.5

1.2 -44

1.7

the dimer [10_NO]₂(as well as [10]₂but not [1]₂), steric bulk of

2

P(t-Bu)₃lowers the activation barrier of PMe₃analog by 20

kcal/mol. However dramatic, this reduction in ∆G_obs^qwith P(t-

Bu)₃apparently falls short of enabling quantifiable Ar-F

reductive elimination in practice.

^aBrønsted R ≈ ∂(∆H^q)/∂(∆H°).

Discussion

DFT calculations show concerted reductive elimination of

Ph-F from 1, a representative of organometallic Pd(II) inter-

mediates featured in catalytic transformations, to be nearly

thermoneutral (∆H° ) +1.2) and face a surmountable activation

barrier, ∆H^q) 28.9 kcal/mol. Thus, Ar-F reductive elimination

in principle is an elementary C-F bond-forming reaction that

can be both kinetically competitive and thermodynamically

feasible in a three-coordinate Pd(II) environment.

Auxiliary ligand effects on elimination kinetics elucidated

computationally can be readily rationalized with help of NBO

partial charges that reveal a monotonic transfer of comparable

0.30 e^-from F^-and 0.35 e^-from Ph mainly to Pd (0.56 e^-) to

accompany Ph-F elimination from 1 with an intermediate

degree of bond-making/breaking and charge-transfer featured

in the transition state. Such pattern of charge redistribution

suggests interactions between Pd and auxiliary donor ligands,

as well as those between F^-and external electrophiles, to be

the strongest in the Pd(II) reactant state, containing most Lewis

acidic Pd and most Lewis basic F^-, and progressively weaken

on going to elimination TS and Pd(0) product. Thus, increasing

the auxiliary ligand donor strength, their number, or binding

instead a hydrogen-bond donor to eliminating F^-in all states

of the reaction profile will all serve to decrease elimination

driving force and increase activation barrier to a necessarily

smaller extent. These qualitative trends can be seen in the

relative activation and equilibrium enthalpies computed for the

complexes grouped in the top two blocks of Table 2.

L ) P(C₆H₄-2-Trip)(t-Bu)₂. Recent studies by Buchwald et

al. demonstrated a range of rather unique catalytic reactivity of

Pd complexes supported by this and related derivatives of P(t-

Bu)₃, including the unprecedented synthesis of phenols,^24hbroad

range of aryl ethers^65a,b,24gand hetero-biaryls^65camong other

applications. Remarkably, both 17 and 18 proceeded to yield,

in separate reactions in the presence of 4 equiv of P(C₆H₄-2-

Trip)(t-Bu)₂in C₆D₆at 60 °C over 22 h, ca. 10% of F-C₆H₄-

1

4-NO₂, readily identified by H, ¹⁹F NMR, and GC-MS in

comparison to authentic sample. This is the first indication, to

our knowledge, of net Ar-F reductive elimination operating to

a quantifiable extent from a transition metal aryl fluoride.

Although a detailed study of Buchwald’s P(C₆H₄-2-Trip)(t-Bu)₂

reactivity with fluoride-bridged Pd dimers is beyond the scope

of this work, experimental observations at hand fit closely within

the mechanistic framework of the original aryl-halide reductive

elimination reactions from Pd(II)²⁵that follow associative

substitution of P(o-Tolyl)₃with bulkier P(t-Bu)₃and cleavage

of halide-bridged dimers into three-coordinate monomers.^25dThe

limited yet measurable success of P(C₆H₄-2-Trip)(t-Bu)₂in

enabling net Ar-F reductive elimination is therefore doubtless

due to its ability to destabilize sterically the fluoride-bridged

Pd dimer beyond the extent possible with P(t-Bu)₃.

(63) Both Pd fragments in the partly cleaved dimer are derived from the basic

T-shaped PdAr(F)L geometry by coordination of a nucleophile to Pd or

electrophile to F, modifications that were computed (Figures 2, 3) to increase

the barrier to reductive elimination from that of the T-shaped precursor.

(64) Underestimation of the stability of _PtBu10_NOrelative to [_PtBu10_NO]₂and

Consideration of Brønsted R parameters (≈ ∂(∆H^q)/∂(∆H°))

for selected subgroups of reactions reveals additional qualitative

aspects of the influence of different modifications to the

coordination environment on the elimination kinetics. Thus,

reduction of auxiliary ligand donor strength from Me₂NHC to

PPh₃destabilizes Pd(II) reactant and stabilizes Pd(0) product

state both to a comparable extent relative to the elimination TS.

Reduction of the activation barrier in this series is a consequence

of increasing driving force with R ) 0.50 fraction of the latter

3

2

3

2

q

_PtBu10-11_NOin the computational model due to inadequate description

3

2

q

of P(t-Bu)CH‚‚‚Pd agostic interactions (vide supra) serves to raise ∆G₁

q

approximated by ∆H°([_PtBu10_NO]_2f1), but affects ∆G_1/2much less due

3

2

to cancellation of the decrease in ∆H°([_PtBu10_NO]

_2f1) and increase in ∆G^q-

3

2

(

_PtBu10-11_NO^q); the difference from experiment introduced by the

3

2

systematic error of the computational model is estimated to be on the order

of 5 kcal/mol.

(65) (a) Burgos, C. H.; Barder, T. E.; Huang, X.; Buchwald, S. L. Angew. Chem.,

Int. Ed. 2006, 45, 4321-4326. (b) Vorogushin, A. V.; Huang, X.; Buchwald,

S. L. J. Am. Chem. Soc. 2005, 127, 8146-8149. (c) Billingsley, K. L.;

Anderson, K. W.; Buchwald, S. L. Angew. Chem., Int. Ed. 2006, 45, 3484-

3488.

9

J. AM. CHEM. SOC. VOL. 129, NO. 5, 2007 1355

A R T I C L E S

Yandulov and Tran

translating into the former. A distinct pattern of influence is

found for the middle block of complexes in Table 2, with

dissociation of strong donor ligands ((MeNHC)CH₂-, PMe₃)

from Pd, of hydrogen bonding between eliminating F^-and HF

and replacement of PMe₃with the highly sterically demanding

P(t-Bu)₃all following a common linear correlation between ∆H^q

and ∆H° with R ) 0.94. All of these modifications of Pd

coordination sphere therefore primarily destabilize Pd(II) reac-

tant state relative to both the elimination TS and Pd(0) product

and serve to lower the elimination barrier directly. Last, variation

of para-substituents of the aromatic ring and replacement of Ph

with a formyl group altogether (bottom block in Table 2)

produce ∆H^q/∆H° correlations with large absolute values of R

(.1) signifying predominant transition state effects. Structural

similarity between TS 1-2^qand a Meisenheimer intermediate

supports S_NAr-type electronic interactions between F^-, Pd, and

the π* orbital of the eliminating fragment as the origin of such

substituents TS effects.

provided the additional steric “push” needed to assist dissocia-

tion of the fluoride-bridged dimer and led to the formation of

measurable quantities of the sought Ar-F. Conversely, L )

P(o-Tolyl)₃analog 17 produced no trace of Ar-F and is

expected to face an even greater barrier to dissociation than

18, based on both the 1.6(3) kcal/mol higher measured barrier

to single bridge cleavage in the synTanti exchange (vide supra)

and the direct observation of coupled equilibria involving

monomeric PdPh(Br)P(t-Bu)₃and dimeric [PdPhP(o-Tolyl)₃(µ-

Br)]₂.^25e

Alternative strategies for promoting Ar-F reductive elimina-

tion could involve the use of judiciously chosen, strong fourth

ligands in coordinatively saturated square-planar environment

stable to formation of fluoride-bridged dimer. Iodide emerged

as a particularly promising direction from calculations (Figure

6), yet experimentally its addition to 17 or 18 resulted in halide

exchange driven by irreversible Hoffman elimination of Bu₄-

NF in C₆D₆and considerably more facile Ar/I^-ligand redis-

tribution⁵⁸of the {PdAr(I)L} products leading to reductive

elimination of biaryls. Still, it is possible that sufficiently inert

π-acidic ligands could indeed produce a stable four-coordinate

Pd(II) environment reactive by Ar-F reductive elimination. At

the same time, use of NHC ligands does not appear to be

immediately promising, e.g., for reasons of greater oxidative

stability, due to the exceedingly facile computed Ar-NHC⁺

elimination from Pd(II) geometry best suited for Ar-F reductive

elimination. Furthermore, the feasibility of F^--mediated in-

tramolecular oxidation of highly sterically demanding PR₃by

Pd(II) has yet to be confirmed. No evidence for formation of

PF₂(t-Bu)₃was obtained from reactivity studies of 18, with C-H

activation, possibly mediated by F, dominating the decomposi-

tion pathways.

All of the identified means to promote Ar-F reductive

elimination kinetics from Pd(II) environment could be poten-

tially applied in practice. The most significant net reduction of

the activation barrier results from destabilization of the square-

planar Pd^IIAr(F)L₂reagent by removal of a strong auxiliary

donor ligand (Figures 2, 6; ∆∆H^q) -22, -14 kcal/mol,

respectively). Exclusion of intermolecular hydrogen-bond donors

(Figure 3, ∆∆H^q) -11 kcal/mol) and increase of steric

repulsion between the eliminating Ar and the sole remaining

auxiliary ligand (Figure 5, ∆∆H^q) -4 kcal/mol) offer

additional means to lower the activation barrier. Last, fine-tuning

improvement of the elimination kinetics should be possible via

lowering of auxiliary ligand donor strength (Figures 1, 5; ∆∆H^q

) -4.8 kcal/mol) and progressively electron-withdrawing

para-substitution of the Ar group (Figures 1, 5; ∆∆H^q) -5

kcal/mol).

Conclusions

Using computational (DFT) methods, aryl-fluoride reductive

elimination is shown to be a feasible elementary C-F bond

forming reaction that can take place from a three-coordinate

Pd^IIAr(F)L environment with activation energy and driving force

generally compatible with synthetic applications. Loss of strong

fourth ligands from Pd or hydrogen bond donors from F each

results in substantial destabilization of Pd(II) reactant, as does

the increase of steric interactions between Ar and L, and yields

most pronounced net reductions of the activation barrier.

Lowering the donor strength of L assists elimination kinetics

additionally by increasing the driving force, while electron-

withdrawing para-substituents on Ar exert a useful S_NAr-like

transition state effect. The key remaining obstacle to effecting

Ar-F reductive elimination in practice was shown to be stability

of fluoride-bridged dimers to complete dissociation into mono-

mers, in an experimental study of the novel [Pd(C₆H₄-4-NO₂)L-

(µ-F)]₂(L ) P(o-Tolyl)₃, 17; P(t-Bu)₃, 18). Inter-ligand steric

repulsion with L ) P(t-Bu)₃provided a 20 kcal/mol destabiliza-

tion of dimer 18, estimated with DFT relative to L ) PMe₃,

however was not sufficient to promote monomer formation to

a significant extent and yield greater than trace amounts of Ar-

F. Ligand (P(t-Bu)₃) C-H activation, possibly assisted by Pd-

bound F^-, followed by isobutylene elimination was the pre-

dominant decomposition mode of 18 observed, while Ar/F^-

scrambling and reductive elimination of Ar-Ar dominate

thermal reactivity of 17. However, use of the bulkier L )

P(C₆H₄-2-Trip)(t-Bu)₂provided the additional steric pressure

However, all of the above strategies for promoting the desired

Ar-F reductive elimination reactivity in practice can be useful

only once measures are taken to prevent aggregation of the

reactive three-coordinate Pd^IIAr(F)L into fluoride-bridged dimer.

Structural preferences in the Pd(Ar)X(Q-Phos) (X ) Cl, Br,

I)^47band Pd(Ar)X(P(t-Bu)₃) (X ) Br, I)²⁵series together with

pronounced stability of dimeric 18 found in this study show

dramatically increased propensity of three-coordinate PdAr(X)L

to dimerization with lighter halides X. ONIOM calculations

provide the equilibrium ∆H° for complete dissociation of the

accurate model of 18, [_PtBu10_NO]₂at 27 kcal/mol, and show

3

2

steric effect of the bulky P(t-Bu)₃to decrease dimer stability

by 20 kcal/mol from that of the PMe₃analog [10_NO]₂(Figure

2

9), primarily via repulsion with the cis-aryl ligand. It is the steric

bulk of the auxiliary ligands that is therefore the key element

of the Pd coordination environment that may be used to

destabilize fluoride-bridged dimer geometry sufficiently for

monomeric PdAr(F)L to become accessible, with only a weak,

agostic stabilization of Pd introduced in place of the fluoride

bridge.

Success of this approach in practice with L ) P(t-Bu)₃in 18

was ultimately limited to formation of ppm levels of the aryl

fluoride product, by ligand C-H activation processes taking

place in part as a consequence of close spatial proximity of the

P(t-Bu)₃hydrogens to bridging fluoride(s). Surprisingly how-

ever, use of Buchwald’s P(C₆H₄-2-Trip)(t-Bu)₂appears to have

9

1356 J. AM. CHEM. SOC. VOL. 129, NO. 5, 2007

Aryl-Fluoride Reductive Elimination from Pd(II)

A R T I C L E S

with BS I, which includes SDD quasirelativistic pseudopotentials on

Pd (28), P (10), Cl (10), Br (28) and I (46 core electrons) with their

associated basis sets (Pd: (8s7p6d)/[6s5p3d];^68a,bP: (4s4p)/[2s2p]; Cl,

Br, I: (4s5p)/[2s3p]^68c) augmented by polarization functions (Pd: f,

1.472;^68dP: d, 0.387; Cl: d, 0.640; Br: d, 0.428; I: d, exponent 0.289

^68e), and 6-31G(d,p)^68fon H, C, N, O, F. This and larger basis sets

employed Cartesian basis functions throughout. The kind of all

stationary points obtained from full optimizations was confirmed via

frequency analysis, which revealed zero and one imaginary frequency

for ground and transition states, respectively, and was used to compute

thermochemical parameters without scaling. To confirm the assigned

nature of the activated complexes, all transition states were animated

according to the normal mode corresponding to the imaginary frequency

and representative structures were optimized to the minima they

connected after perturbing the TS geometry. Detailed in Supporting

Information, partial IRC calculations were used to additionally char-

acterize the TS for reductive elimination of Ph-F from PdPh(F)(Me₂-

NHC), while a series of constrained optimizations was used to explore

in greater detail reductive elimination of [Me₃P-Ph]⁺from PdPh(F)-

(PMe₃). For all BS I-optimized geometries, single-point energy calcula-

tions were performed with BS II, in which SDD pseudopotentials are

used for Pd (28), Br (28) and I (46 core electrons) with the associated

SDD basis set augmented by (2f1g) functions^68gfor Pd and Martin’s

SDB-cc-pVTZ basis sets used for Br and I,^68gwith the other atoms

described by 6-311G+(2d,p) basis sets.^68h-jElectronic energies from

the BS II calculations were combined with thermochemical corrections,

computed via BS I frequency analyses at STP without scaling, to

produce the improved estimates of gas phase ∆H° and ∆G° presented

in the Figures and text exclusively, unless noted otherwise. Solvation

energies were computed for selected geometries with BS I in C₆H₆

solvent described with a conductor-like screening model CPCM⁶⁹and

UAKS atomic radii (electrostatic contributions only). Benchmark study

of (o-Tolyl)-Br reductive elimination from Pd((o-Tolyl))Br(P(t-Bu)₃)

additionally included single-point calculations on BS I-optimized

geometries with BS III: BS II with Cl and lighter atoms described by

AUG-cc-pVTZ basis sets,^68k-musing loose SCF convergence criteria.

NBO analysis⁷⁰was done with BS I. Basis set superposition errors

(BSSE) for halide-bridged dimers were computed with counterpoise

procedure⁷¹and BS I individually, recomputed for selected dimers with

BS II and extrapolated from BS I to BS II for the other, closely related

structures from the computed BS II/BS I ratios chosen as benchmarks;

final BS II values of BSSE for all dimers analyzed were between 1.4

and 1.7 kcal/mol.

needed for dimer dissociation and resulted in formation of aryl-

fluoride net reductive elimination product in low yields (10%)

in reactions with both 17 and 18.

Experimental Section

General Methods. All manipulations of air- and moisture-sensitive

compounds were carried out by standard Schlenk and glovebox

techniques under atmosphere of nitrogen using flame- and oven-dried

glassware, including NMR tubes and inserts. Full details are provided

in Supporting Information.

[Pd(C₆H₄-4-NO₂)P(o-Tolyl)₃(µ-F)]₂(17). A mixture of [Pd(C₆H₄-

47

4-NO₂)P(o-Tolyl)₃(µ-I)]₂(880 mg, 0.667 mmol), 4 equiv of AgF

(338.4 mg, 2.667 mmol), and 0.1 equiv of P(o-Tolyl)₃(20.3 mg, 66.7

µmol) was sonicated in PhMe (20 mL) at 25 °C in the dark for 6 h, at

which time NMR of an aliquot in CDCl₃showed complete conversion

of the starting iodide. The rusty-brown solid was filtered off with Celite,

the filtrate was brought to dryness in vacuo, and the residue evacuated

for overnight at RT. Trituration with 40 mL of Et₂O for several hrs at

RT afforded light mustard microcrystalline solid, which was collected

on a frit, washed with Et₂O at RT and dried in vacuo at +70 °C for 6

h. Yield 0.632 g (0.573 mmol, 86%), mixture of anti- and syn-isomers.

¹H NMR (C₆D₆, 22 °C): δ 7.63 (br, 4H, (C₆H₄-4-NO₂)₂), 7.47 (d, J_HH

) 8.8 Hz, 4H, (C₆H₄-4-NO₂)₂), 6.95 (m, 24 × 0.38 H, P(C₆H₄-2-Me)₃,

syn), 6.77 (m, 24 × 0.62 H, P(C₆H₄-2-Me)₃, anti), 2.40 (s, 18 × 0.62

H, P(C₆H₄-2-CH₃)₃, anti), 2.29 (s, 18 × 0.38 H, P(C₆H₄-2-CH₃)₃, syn).

³¹P{¹H} NMR (C₆D₆, 22 °C): δ 32.8 (d, J_PF) 167 Hz, 2 × 0.6 P,

P(o-Tolyl)₃, anti), 31.9 (d, J_PF) 167 Hz, 2 × 0.4 P, P(o-Tolyl)₃, syn).

¹⁹F NMR (C₆D₆, 22 °C): δ -292.7 (td, J_PF) 167 Hz, J_FF) 77 Hz, 1

× 0.4 F, trans-(o-Tolyl)₃P-Pd-F, syn), -293 (obscured, 2 × 0.6 F,

PdF, anti), -314 (br, 1 × 0.4 F, cis-(o-Tolyl)₃P-Pd-F, syn). Anal.

found (calcd., %) for C₅₄H₅₀F₂N₂O₄P₂Pd₂: C 58.51 (58.76), H 4.53

(4.57), N 2.48 (2.54).

[Pd(C₆H₄-4-NO₂)P(t-Bu)₃(µ-F)]₂(18). A mixture of [Pd(C₆H₄-4-

NO₂)P(o-Tolyl)₃(µ-F)]₂(400 mg, 0.362 mmol) and 2.2 equiv of P(t-

Bu)₃(161.3 mg, 0.797 mmol) was stirred in PhMe (20 mL) for an

hour at 25 °C, filtered through Celite, the filtrate was brought to dryness

in vacuo, and the residue dried at +55 °C for 7 h. Trituration with 20

mL of Et₂O for several hrs at RT afforded pale-mustard microcrystalline

solid, which was collected on a frit, washed with Et₂O at RT, and dried

in vacuo at +65 °C for 4 h. Yield 0.174 g (0.193 mmol, 53%), mixture

1

of anti- and syn-isomers. H NMR (C₆D₆, 22 °C): δ 7.77 (br, 4H,

(C₆H₄-4-NO₂)₂), 7.51 (s, 4H, (C₆H₄-4-NO₂)₂), 1.14 (d, J_PH) 12.8 Hz,

54 H, P(C(CH₃)₃)₃). ³¹P{¹H} NMR (C₆D₆, 22 °C): δ 80.2 (second order

ONIOM calculations⁷²were performed with a two-layer scheme, in

which PMe₃-based QM/MM model system was defined within P(t-

Bu)₃-based MM real system; the difference between real and model

systems in all complexes studied was confined to the phosphine ligands.

m, br, J_PF(trans) ) (154 Hz, J_PF(cis) ) - 10.7 Hz, 2 P, P(t-Bu)₃). ¹⁹

F

NMR (C₆D₆, 22 °C): δ -286.9 (second order m, J_PF(trans) ) (154

Hz, J_PF(cis) ) - 10.7 Hz, |J_FF| ) 119 Hz, 2 F, PdF). Anal. found

(calcd., %) for C₃₆H₆₂F₂N₂O₄P₂Pd₂: C 48.51 (48.06), H 7.31 (6.95), N

2.87 (3.11). X-ray quality crystals were grown from PhF solution

at -35 °C.

Thermolysis Studies. Flame-dried J. Young tubes with or without

Teflon inserts and an internal standard in the form of CF₃COOH/(CF₃-

CO)₂O sealed in a glass capillary were used in a temperature-controlled

oil bath, with periodic monitoring by NMR (¹H, ³¹P{¹H} and ¹⁹F) at

ambient T °C. Products were analyzed and identified on the basis of

NMR, GC-MS, and ESI-MS, following basic workup as necessary. See

Supporting Information for characterization of new products and

additional details.

(68) (a) Andrae, D.; Ha¨ussermann, U.; Dolg, M.; Stoll, H.; Preuss, H. Theor.

Chim. Acta 1990, 77, 123-141. (b) Andrae, D.; Ha¨ussermann, U.; Dolg,

M.; Stoll, H.; Preuss, H. Theor. Chim. Acta 1991, 78, 247-266. (c) Bergner,

A.; Dolg, M.; Ku¨chle, W.; Stoll, H.; Preuss, H. Mol. Phys. 1993, 30, 1431.

(d) Ehlers, A. W.; Bo¨hme, M.; Dapprich, S.; Gobbi, A.; Ho¨llwarth, A.;

Jonas, V.; Ko¨hler, K. F.; Stegmann, R.; Veldkamp, A.; Frenking, G. Chem.

Phys. Lett. 1993, 208, 111-114. (e) Ho¨llwarth, A.; Bo¨hme, M.; Dapprich,

S.; Ehlers, A. W.; Gobbi, A.; Jonas, V.; Ko¨hler, K. F.; Stegmann, R.;

Veldkamp, A.; Frenking, G. Chem. Phys. Lett. 1993, 208, 237-240. (f)

Hariharan, P. C.; Pople, J. A. Theor. Chim. Acta 1973, 28, 213. (g)

Martin, J. M. L.; Sundermann, A. J. Chem. Phys. 2001, 114, 3408-3420.

(h) Krishnan, R.; Binkley, J. S.; Seeger, R.; Pople, J. A. J. Chem. Phys.

1980, 72, 650. (i) McLean, A. D.; Chandler, G. S. J. Chem. Phys. 1980,

72, 5639. (j) Frisch, M. J.; Pople, J. A.; Binkley, J. S. J. Chem. Phys. 1984,

80, 3265. (k) Woon, D. E.; Dunning, T. H., Jr. J. Chem. Phys. 1993, 98,

1358. (l) Kendall, R. A.; Dunning, T. H., Jr.; Harrison, R. J. J. Chem.

Phys. 1992, 96, 6796. (m) Dunning, T. H., Jr. J. Chem. Phys. 1989, 90,

1007.

Computational Details. All calculations were carried out with

Gaussian 03 suite of programs,⁶⁶using hybrid density functional method

B3PW91⁶⁷for QM models. No symmetry constraints were imposed

throughout with a single exception of [PdPh(Me₂NHC)(µ-Cl)]₂(C₂).

All QM geometries were optimized under standard convergence criteria

(69) (a) Barone, V.; Cossi, M. J. Phys. Chem. A 1998, 102, 1995. (b) Cossi,

M.; Rega, N.; Scalmani, G.; Barone, V. J. Comput. Chem. 2003, 24, 669.

(70) (a) Glendening, E. D.; Reed, A. E.; Carpenter, J. E.; Weinhold, F. NBO,

Version 3.1. (b) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem. ReV. 1988,

88, 899-926.

(71) (a) Simon, S.; Duran, M.; Dannenberg, J. J. J. Chem. Phys. 1996, 105,

11024. (b) Boys, S. F.; Bernardi, F. Mol. Phys. 1970, 19, 553.

(66) Pople, J. A. et al. Gaussian 03, revision C.02; Gaussian, Inc.: Wallingford,

CT, 2004.

(67) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (b) Perdew, J. P.; Wang,

Y. Phys. ReV. B 1992, 82, 284.

9

J. AM. CHEM. SOC. VOL. 129, NO. 5, 2007 1357

A R T I C L E S

Yandulov and Tran

QM part was described at the B3PW91/BS I level and MM contribu-

tions were computed with UFF. Factors defining distances to link atoms

in model system at high and low levels were determined from UFF

(model and real) and B3PW91/BS I (model system) geometry optimiza-

tions in a representative case. Loose convergence criteria were used in

problematic transition states optimizations (a rationale is provided in

Supporting Information); for comparison with values obtained with pure

QM models, we estimate the energy difference that resulted from using

loose criteria to be under 0.1 kcal/mol. As for pure QM models, both

ground and transition states optimized with ONIOM were subjected to

frequency analysis that revealed the requisite zero and one imaginary

frequency for the respective kinds of stationary points. Improved

estimates of gas phase ∆H° and ∆G° were obtained likewise, from

ONIOM/BS II//ONIOM/BS I single point calculations, while CPCM

solvation energies were evaluated with B3PW91/BS I on the real system

(ONIOM/BS I) geometry. BSSE’s were computed with extracted QM

parts.

Acknowledgment. This work was partially supported by the

National Center for Supercomputing Applications under a

developmental allocation and utilized IBM P690. We thank Dr.

Dmitry V. Khoroshun for advice on ONIOM optimizations, Drs.

Frederick J. Hollander and Allen G. Oliver of the UC Berkeley

College of Chemistry X-ray Diffraction Facility (CHEXRAY)

for collection of X-ray diffraction data, and Stanford University

for funding.

Supporting Information Available: Complete reference 66,

computational data; variable-temperature NMR data and details

of line shape analysis; crystal data and structure refinement,

atomic coordinates and equivalent isotropic displacement pa-

rameters, bond lengths and angles, and anisotropic displacement

parameters for [Pd(C₆H₄-4-NO₂)P(t-Bu)₃(µ-F)]₂×(PhF)₂; and

experimental details of characterization of thermolysis products

of 17 and 18. This material is available free of charge via the

Internet at http://pubs.acs.org.

(72) (a) Maseras, F.; Morokuma, K. J. Comput. Chem. 1995, 16, 1170-1179.

(b) Svensson, M.; Humbel, S.; Froese, R. D. J.; Matsubara, T.; Sieber, S.;

Morokuma, K. J. Phys. Chem. 1996, 100, 19357-19363. (c) Vreven, T.;

Byun, K. S.; Koma´romi, I.; Dapprich, S.; Montgomery, J. A., Jr.;

Morokuma, K.; Frisch, M. J. J. Chem. Theory Comput. 2006, 2, 815-826.

JA066930L

9

1358 J. AM. CHEM. SOC. VOL. 129, NO. 5, 2007

Article Doi

Aryl-fluoride reductive elimination from Pd(II): Feasibility assessment from theory and experiment

DOI: 10.1021/ja066930l

Source and publish data:

Authors:

Article abstract of DOI:10.1021/ja066930l

Full text of DOI:10.1021/ja066930l

Products guided by the article

R&D Labs maybe for 929110-40-7

Relevant to this article

Hot Product