Literature Seminar (D2 part) 071114 (wed) Wataru Itano (D2) page 1

Cationic Phosphine-Gold(I) Catalysis for C-C Forming Reaction with C C ~ Can relativistic effects rationalize its reactivity? ~

@ Carbometallation is powerful C-C bond formation although stoichiometric metal is needed.

Nu M Nu M E Nu E

(M = Li, MgX, Cu, Zr etc.)

outstanding example LHMDS. THF/tol 1/1, -78°C; tBuLi (4 eq); AcOH A.G. Myers et al. 30-40% (JACS 2006 14825)

@ Catalytic variant of this transformation is highly attractive

cat. M Nu E Nu E

Electorophilic late transion-metals (easily forming π-complex) might be suitable for Nu addition.

But standard carbon Nu (enolate etc.) is difficult to apply (oxi./red. stable but electrophic metal is desirable) problem : oxidative coupling of substrate etc.

Recently some transition metal catalysts (InIII, AuI or III, PdII etc) are realized for this carbon Nu addition. I Among them, [R3P-Au ] is one of the most powerful catalysts.

-1995 B.Sc. and M.Sc. / University of Toronto, Canada (Prof. Ian W. J. Still) 1995-2000 Ph.D. / Stanford University (Prof. Barry M. Trost) · asymmetric alkylation (Pd), cycloisomerization (CpRu+), some total syntheses · 21 papers (17 JACS ! ) 2001-2002 Post-Doctoral Fellow / Caltech (Prof. Robert H. Grubbs)

2002-2006 Assistant Professor / University of California, Berkeley 2006- Associate Professor / University of California, Berkeley F. Dean Toste

He utilizes relativistic effect to explain unique reactivities of gold.

?（相対論的効果） ? D. J. Gorin and F. D. Toste (Nature 2007, 446, 395-403)

< Contents > 0. Preface (p2) 3. Reaction after Nu addition (via carbenoid) 1. Thoretical chemistry of Gold 3-1. Generation of cationic intermediates (p9) (p10) 1-1. Relativistic effects (p3) 3-2. Investigation of cationic intermediate 1-2. Other theoretical aspects (p4)

2. Nu addition toward C=C + 2-1. Heteroatom Nu (discovery of [R3PAu(I)] ) (p5) 2-2. Theoretical chemistry of cationic phosphineGold(I) (p6) 2-3. Carbon Nu (Au(I) vs Au(III)) (p7-8) Group 11 0. Preface page 2 14 10 1 Au --- [Xe] 4f 5d 6s Cu > Au Ag I 14 10 0 (Z = 79) (Au --- [Xe] 4f 5d 6s ) Period 6 Ir Pt Au Hg L (π-acceptor) ? L (σ-donator) I X Au L AuI X 14e AuIII L AuIII 16e (d10) linear, two-coordinate (d8) X X planar, tetra-coordinate > intrinsically π-electrophilic

first recognized in heterogeneous catalyst

G. J. Hutchings (J. Cat. 1985, 96, 292.)

(H-OR, H-NR, H-Ar) then extend homogeneous catalyst

O O (Na[AuCl4], AuCl3, AuCl) Nu = R OR (Toste`s initial work on Gold) [Au] = cationic phosphine-gold(I)

(His understanding from experiments)

1. strong Lewis acidity (compared with other π-electrophilic late transition-metals) R P AuI X 3 2. potential to stabilize cationic intermediate

> Relativistic effect (量子化学の世界において相対性理論を加味した場合に出てくる効果…らしい） should be considered when investigating (calculating) the electronic structure of heavy atoms (> 5th-row)

(stabilized energically) LUMO in Au(I) In Au (Z = 79), 1. contracted 6s orbitals relativistically expected explains these characters. 2. expanded 5d orbitals

(destabilized energically)

HOMO in Au(I) > To simplify Toste`s work

(with alkyne) E Nu direct addition of carbon Nu to alkyne E "alkynophilic acid" Au 4 (see section 2) Au Nu E Nu Au Nu 1 Au 23 E 5 Au (functionalization relativistic 6s contraction E (hopefully) of carbenoid) explains Lewis acidity product generation of further intermediate; Nu Au functionalization (see section 3) E Au relativistic 5d expansion 6 might stabilize further intermediate (Au-carbenoid ?) 1. Theoritical chemistry of Gold page 3 1-1 Relativistic effects > quantum mechanics + special relativity Schrodinger`s equation molecular orbital calculation structure of electrons (speed of light : c is infinite)

· the average radial v of the 1s electrons

v = Z au In heavy atom (Z > 4-50), v has significant value relative to c Z: atomic number (c = 137 au) consideration Dirac`s equation of special relativity (c is finite)

The term "relativistic effects" refers to any phenomenon resulting from the need to consider velocity as significant relative to the speed of light.

> one basic consequense of special relativity m m: corrected mass m0: non-relativistic (rest) mass mass increases towards infinity as a body's v approaches c

> effects Effect 1: The Relativistic Contraction mが相対論的効果で増加 Bohr radius: a0は減少

(supported by caclulation) (expected r for valence s) Cu(4s) Ag(5s) Au(6s) r [au] 3.2 3.4 3 P. Pyykko et al. (Acc. Chem. Res. 1979, 276.)

P. S. BAGUS (Chem. Phys. Lett. 1975, 408.)

· contraction of the 1s and all s and p orbitals. why Au so contracted around Z ~80 ?

78Pt Hg Stabilization of orbital energies 80

Practically, contraction is only significant for elements in which the 4f and 5d orbitals are filled

· electrons are closer to the nucleus; have greater ionization energies. (supported by caclulation)

HF (non relativistic) DHF (relativistic) ps (non lanthanide)

(experimental values) Effect 2: The Relativistic Self-Consistent Expansion page 4

The d and f orbitals are not contracted. (higher angular momentum, seldom descend to nucleus) better shielded by contracted s and p orbitals Instead, see a weaker nuclear attraction

Expansion of d and f orbitals Destabilization of orbital energies

Effect 3: The Spin-Orbit splitting

relativistic effects 6s orbitals contracted stabilizing Au 5d orbitals expanded destabilizing (and spliting)

> these effects are reflected in MO calculation P. Pyykko et al. (Acc. Chem. Res. 1979, 276.)

1-2 Other theoretical aspects > aurophilicity the tendency for Au–Au interactions to be stabilizing on the order of hydrogen bonds (might be important for neutral complex (LAgX etc..))

I > [Me2Au ] not particularly nucleophilic relative to the corresponding CuI (and AgI) complexes. E. Nakamura et al. (J. Am. Chem. Soc. 2005, 127, 1446.) III > Me2RAu L (R = Me or allyl, L = PMe3) reductive elimination is relatively disfavoured as well.

AuI and AuIII complexes do not readily cycle between oxidation states. (of course exception exists) ( + from experiments) tolerate both oxygen and acidic protons,

@ With its unique properties strongly influenced by relativistic effects, theoretical chemists have much attention to gold.

review P. Pyykko (Angew. Chem. Int. Ed. 2004, 43, 4412.)

+ + (pp 4424) one can regard the moiety [AuPR3] as a σ-acceptor in analogy to H . 2. Nu addition toward C C page 5 A. S. K. Hashmi (Angew. Chem. Int. Ed. 2005, 44, 6990-6993) A. S. K. Hashmi and G. J. Hutchings (Angew. Chem. Int. Ed. 2006, 45, 7896-7936) review quite a lot ! A. S. K. Hashmi (Chem. Rev. 2007, 107, 3180-3211) D. J. Gorin and F. D. Toste (Nature 2007, 446, 395-403)

2-1 Heteroatom Nu III > initially Au halide (intermolecular) (intramolecular) (anhydrous)

6-exo-dig

II @ Pd unsatisfactory results @ internal --- regio mixture, terminal --- Markovnikov

Utimoto. K. et al. (Heterocycle. 1987, 96, 292.) @ K[Au(CN)2] is inactive @ Na[AuCl4] quickly reduced to metalic Au (TON >50) Utimoto. K. et al. (J. Org. Chem. 1991, 96, 292.) > cationic phosphineAuI

J. H. Teles et al. (Angew. Chem. Int. Ed. 1998, 37, 1415)

in situ generation of [L-AuI]+

Ph PAuMe (X mol%) 3 < terminal alkyne > Me H + MeOH MeSO3H (10X mol%) MeO OMe less steric hinderance for Au (Markovnikov) (continuous streaming) (1 eq) 20-50 °C

TON up to 50000! (0.00002 mol%), TOF up to 5400 h-1 (cf HgII quickly reduced to metalic Hg (TON ~100))

I + -1 [L-Au ] the initial TOFs [h ] : Ph3As (430) < Et3P (550) < Ph3P (610) < (4-F-C6H4)3P (640) < (MeO)3P (1200) < (PhO)3P (1500) electron-poor ligands lead to an increase in activity, but the stability decreases.

I -1 - - -3 - -3 [Ph3PAu X] the initial TOFs [h ] : I (2) < Cl (7) < NO ~ CF3COO ~ CH3SO (700) progresses from soft to hard anions

+ < internal alkyne > less steric hinderance for MeOH

@ Ab initio calculation @ experimental data

cis-auration

intuitively trans-auration acceptable

1,3 hydrogen migration; bond rotation

theoretically cis-auration 2-2 Theoretical chemistry of cationic phosphineGold(I) page 6 I + > [R3PAu ] + PH3 far more covalent robust + 1) Au(I) is a large, diffuse cation sharing positive charge with the phosphine ligand, I [H3P Au ] in reaction one might expect orbital rather than charge interactions 2nd PH is 3 2) in binding a second ligand. PH3 "soft" I + 6s orbital of [R3PAu ] can further accept electrons (contracted 6s) keep strong Lewis acidity

AuCl III + + PH3 not investigated [X2Au ] P. Schwerdtfeger et al. (Inorg. Chem. 2003, 42, 1334.)

I + > Stability of [R3PAu ] + L 1) with several ligand stability [kJ/mol]: J. H. Teles et al. (Angew. Chem. Int. Ed. 1998, 37, 1415) [Me PAuI]+ + L 3 CH2Cl2 (+63) < H2O (+44) < acetylene (+38) < MeOH ~ 1,4-dioxane (+24) < propyne (+18) < THF (+2) < 2-butyne (0) < Me2S (-18 ) < Ph3P (- 114) substituted alkynes are better ligands than MeOH or dioxane carbophilic rather than oxophilic

W. Koch et al. (J. Phys. Chem. 1996, 100, 12253) 2) alkyne vs alkene V. M. Rayon et al. (J. Phys. Chem. A 2004, 108, 3134) Experimentally, reactivity of Nu is but theoretical stability is inverse + + > E(Au -alkyne) < E(Au -alkene) (~10 kcal/mol)

alkynes LUMOs intrinsically lower than alkenes (~0.5eV) That is important for Nu addition.

L L I + > Backdonation from [R3PAu ]

1) to alkyne /alkene (by calculation) Backbonding to antibinding orbital (of alkene / alkyne) is poor (smaller than Cu) render alkyne / alkene more electron deficient

2) to carbene (their explanation, not by calculation ?) Backbonding to antibinding orbital also might be poor but backbonding (from AuI 5d) to lower-energy (than antibonding) non-bonding p-orbitals (of carbene) might be suitable relativistically expanding (destabilizing) This can stabilize carbenoid intermediate in the AuI catalyzed reaction 2-3 carbon Nu page 7 > (Nu = β-keto ester)

(J. Am. Chem. Soc. 2004, 126, 4526)

5-exo-dig

terminal alkyne only

#1 [AgI]+ weak Lewis acidity cationic - AuI - mono-phosphine I + #2 [R3PAg ] insufficient III #3 [Au X3] sufficient Lewis acidity (?) but undesired path ? Au vs Ag --- relativistic contraction of s orbital I I + I #4 [R3PAu X ] insufficient ? [Au ] vs Au --- aurophilicity (?) I III I + ? Au vs Au --- unknown (for regio , see (Nu = Ar-H)) #5 [R3PAu ] sufficient Lewis acidity I + #6 [(RNC)2Au ] insufficient ? Au vs Pt --- not mentioned (possible?)

nucleophilic attack on a Au(I)-alkyne complex by enol trans-auration

cis-auration formation of a Au-enolate followed by a cis-carboauration

D H [Au] [Au] D H

Trans-auration Now generally accepted. (mechanism A)

Me Me MeO C MeO2C 2 [AuI]+ 5-endo-dig (Angew. Chem. Int. Ed. 2004, 43, 5350) OH O internal alkyne OK 5-endo-dig (1,5-enyne) R R

I I > usually prepared in situ R3PAu Cl + Ag X

I R PAuIMe + HX [R3PAu ] X 3

(X = OTf, SbF6, BF4 etc) I (L = R3P) [(R3PAu )3O]X+ HX > (Nu = Ar-H) page 8 @ Regioselectivity is inversed between Au(I) and Au (III) with terminal alkyne, (only intrinsical regio ?) M. T. Reertz et al. (Eur. J. Org. Chem. 2003, 3485)

n+ [Au(III)] [LAu(I)]+ (10 eq) (10 eq)

+

(Z/E = 100/0)

Ph3PAuCl/AgBF4 ineffective AuCl3 / 3AgBF4 (5 mol%) 37 (~60%) + 39 (10%) 60°C, 16h (Z/E = 78/22) Au(III) slightly activate carbonyl ?

@ Ar-[Au] might be unlikely instead of electrophilic activation of alkyne (2 eq)

[AuCl3]2 2,6-lutidine Ph H

THF 50°C, 5h 82-94% unstable

Y. Fuchita et al. (J. Chem. Soc., Dalton. Trans. 2001, 2330) > AuI vs AuIII (Nu = -C=O) @ They said "oxophilic AuIII species" and "π-philic AuI species"

(or AuCl3 + THF)

[Au(III)] [LAu(I)]X 1,2 hydride shift via gold carbenoid H

V. Gevorgyan et al. (J. Am. Chem. Soc. 2005, 127, 10500)

(my opinion) Basically both Au(I) and Au(III) are π-philic But Au(III) is slightly more oxophilic than Au (I)

(summary so far)

Nu E Nu E Nu This section (Nu = N, O, C, E = H) 2 Au Au 3 4 E

a lot ! How to generate same (Nu = S, E = allyl, silyl etc.) vinyl-Au E position Nu Y. Yamamoto et al. Nu (Angew. Int. 2006, 45, 4473.) (endo/exo) E (OL. 2007, 9, 4081.) (R1 = E) Au R1 8 several 9 electrophiles R2 Nu R R (Nu = O, E = benzyl, iminium) Nu R Nu same Nu L. Zhang E (JACS 2005, 127, 16804) (R2 = E) E position (endo/exo) F. D. Toste et al. 7 Au E intramolecular carboalkoxylation Au Au 10 11 12 (JACS. 2006, 128, 12062)

different position Nu Nu Nu E Next section new reaction ? Au E Au (Toste`s plan) 3 E 5 Au 6 relativistically expanding 5d orbitals of Au might stabilize cationic intermediate by backdonation 3. Reaction after Nu addition (via carbenoid) page 9 different position

Nu Nu Nu E Nu E Au E Au 2 Au 3 E 5 Au 6

3-1 Generation of cationic intermediates

@ AuI-NHC complexes are known can catalyze cyclopropanation of styrene with ethyldiazoacetate S. P. Nolan et al. (Angew. Chem. Int. Ed. 2005, 44, 5284)

@ vinyl Au species was proposed to have significant Au carbene character on the basis of NMR, X-ray, and calculations

H. G. Raubenheimer et al. (Organometallics 2002, 21, 3173) (E ---> Cr(CO) ) 5 cationic intermediate [5 6] might exists

different position

E 1,2- E Nu E Nu Nu E Nu H-migration E Nu endo H H H H 17 13 Au 14 Au 15Au 16 Au Au

@ initial work of Toste same intermediates (14-16) was proposed with AuI Nieto-Oberhuber, C. et al. (Angew. Chem. Int. Ed. 2004, 43, 2402) (J. Am. Chem. Soc. 2004, 126, 10858) (Chem. Eur. J. 2006, 12, 1677.) Furstner, A. et al. (J. Am. Chem. Soc. 2004, 126, 8654)

complete conversion, but substantial amount of decomposition. · 5% AuCl3 with 15% AgOTf

pseudoeqatrial

(14 15) (16)

·Au might lower the barrier to cationic intermediate by backbonding. "expanded 5d" stereospecific

Anyway... carbenoid or cationic ? ·With Au(III), analogous intermediates were proposed. ·Pt(II) shows similar reactivity(relativistic effect also influense Pt) cationic Au(I) --- even in the presence of strongly donating ligands, such as phosphines activation of alkynes Pt(II) --- simple salts or CO complex reactivity tuning, stereoselectivity 3-2 Investigation of cationic intermediate page 10 different position

R R R R Nu Nu Nu Nu O O O 5-exo O C-C bond formation E E E E Au Au Au Au 18 19 20 21

(J. Am. Chem. Soc. 2005, 127, 5802)

Original (J. Org. Chem. 1984, 49, 950) Rautenstrauch rearr.

metal carbene intermediate proposed "a method for generation of metal-carbenoid"

Toste`s case chirality transfer not Au-carbene (21)

A. R. de Lera (J. Am. Chem. Soc. 2006, 128, 2434) = (19) DFT calculation predicts helically chiral intermediate

Nazarov like cyclization from cationic vinyl-Au (9 or 11) (20) = cyclization might predominate

carbenoid or cationic ? cationic

= (19) on a fast time scale (intramolecular reaction)

= (21)

toward highly carbenoid character

(J. Am. Chem. Soc. 2005, 127, 18002) cis favored

1)

stereochemical information was not conserved. "carbenoid" might be concerted mechanism page 11 R 2) PivO R'

Au 3) L Ar'

Ar' R PivO high ee even linear geometry of AuI R' Ar X Au Au bifunctional ? Anyway... Au Au carbenoid character might dominates L L L L further reaction would be developed. mono cationic ? dicationic ?

> to conquar demerit of linear coodination geometry of AuI for ee

(Science. 2007, 317, 496)

5-exo

@ standard chiral phosphine poor ee (<10%ee)

@ only PPh3AuCl or Ag-6 or H : no rxn @ dinuclear AuI better @ less polar solvent best (+H+: known as chiral Bronsted acid catalyst) ee depends on the proximity of the counteranion to the cationic gold center

@ match-mismatch pairing between chiral phosphine and chiral counteranion observed in other systems.

might extend standard transition metal catalysed process

cf) hydroamination of allene with chiral phosphine gold (J. Am. Chem. Soc. 2007, 129, 2452)

6s contraction strong Lewis acidity AuI (Nature 2007, 446, 395-403) < summary > 5d expansion backdonation for carbenoid

(with alkyne) carbon Nu Conia-Ene (JACS 2004, 126, 4526) (Angew 2004, 43, 5350) Silyl-enol (Angew 2006, 45, 5991)

How to generate E Nu propalgyl claisen. (JACS 2004, 126, 15978) carbon E ring expansition (JACS 2005, 127, 9708) E alkoxycarbolation (JACS 2006, 128, 12062) etc. Au 4

Nu How to generate Nu E acetylenic Schmidt (JACS 2005, 127, 11260) Au cyclopropanation (JACS 2005, 127, 18002) Au 2 3 1,2-shift, cyclization Nu Nu rearrangement etc.. How to proceed different mode cycloisomerization (JACS 2004, 126, 10858) product Rautenstrauch rearr.(JACS 2005, 127, 5802) E Au E E 5 Au 6 oxidative trap (JACS 2007, 129, 5838)

(with allene) other works about Au (cycloaddition) (asymmetric variants)